No complaints about this particular post (I've not read it), but I always prefer that ACX (and most other blogs) not do guest posts. This applies to the book reviews too, which I never read. These are just spam to me. I'd rather a blog just link to posts which they think readers might be interested in, in one of their other posts (e.g. link post or open thread post), if they want to share them.
Not directly, but they give me money for publishing a blog, I try to publish ~2-3 posts/week to justify this, and if someone else gives me one then I only have to write 1-2.
This feels like a good time to mention that I have a paid subscription because it's something like "investing a world where Scott Alexander is doing well", mostly as gratitude for your old stuff; I don't perceive this as anything like buying a service from you. Of course not everyone is like this, but I thought I'd mention that some of us are.
If you just want to reward Scott for his past contributions and are not terribly interested in the ongoing service he provides by continuing this blog, wouldn’t it be better just to make a direct contribution to him and bypass the whole issue?
Your form of subscription is paying him to write, but with a time delay. If he wants more subscribers like you, Scott should write more, and then wait for them to be grateful.
But (as a paid subscriber), if you model us as just immediately archiving posts not written by you, the computation of whether it's worth it to continue subscribing is downstream of that deletion. So the guest posts don't bring any free money, but they might dupe you into thinking you're delivering a product of the quality you've pre-decided maximizes your audience retention utility.
There are too many “yous” in this comment for me to be completely clear about what you’re getting at, (in the first instance, I think it refers to Scott; in the second, I think it refers to his subscribers) but either way Scott is relying on his credibility so I don’t think it move the needle much. As a further note, I think the OP was kind of needling Scott to defend his credibility, and I think he dodged that nicely.
> There are too many “yous” in this comment for me to be completely clear about what you’re getting at, (in the first instance, I think it refers to Scott; in the second, I think it refers to his subscribers)
They all refer to Scott. The subscribers are referred to as "us". Subscribers obviously do not have a model of audience retention.
The post originally started with a paragraph explaining that the first part was mostly to cast appropriate doubt on my finding, and was unnecessary "if you want the prize without its price." Scott replaced that with his own disclaimer which is clearer/better, but didn't include that.
I'm not sure whether you're being serious here, but FWIW, I don't consider guest posts to be part of what I'm paying for. I'm perfectly happy that the content you produce is good value for what I pay, but it's only that content I'm paying for.
For what it's worth (likely not much), I have explicitly avoided buying a paid subscription because you (and guest posters) write too much for me to keep up with everything.
I try to read all of it, because every once in a while there's a post which ends up permanently forming a fundamental piece of my understanding of something important. (The Nietzsche posts are the most recent example of this.) But when you say "nothing too important is ever going to be behind a paywall" I choose to happily take your word for it.
(Case in point: I have not yet read the post I am currently commenting on. I was scrolling to the comments to see whether this was worth reading in full. I'll probably get to it at some point.)
As others have mentioned - I would not unsubscribe or anything with fewer guest posts/book reviews. I would be perfectly happy with 12 Scott posts a year.
I too wish posts like this were limited to a roundup or separate feed. Like - the book reviews are a lot, but I’m here for when you talk culture wars and politics and medicine and FDA and parenting, and this post and many book reviews… well, they aren’t written by you! There’s no micro humor or charts or research. Just like Reddit and lesswrong posts where I feel like it’s too neurotic and there isn’t a grounding in feelings, emotions, and what normies actually do/think. Scott - you somehow always are both technical but also relate things back to practical real life.
I would rather you post works from lorein as you write them - the posts about adhd, depression, etc would have great discussion.
And I know it’s a bit parasocial relationship of me, but I’d love low content posts as subscriber only stuff. What did you do this week. What toys have your kids been enjoying. What’s on the radio. What car are you thinking about buying.
Everything its opposite! I'm a cat person and have no interest in what Jonah Goldberg thinks but I used to enjoy encountering his twitter feed just for the relentless 10-dog-things- (with occasional cat e.g. taking dog's spot on the couch: "Be the better quadruped" hee hee) to-one-politics thing ratio. Plus he seemed to live in a cool autumnal landscape.
I haven't been able to view twitter in a long time though.
Yes!! I liked that! Overall your writing was interesting and I read it all. I didn’t like the bits about sequences/thinking because that’s not why I’m here, but I’m sure a ton of people did!
I’ve tried to write like Scott in a few long Reddit posts about sunscreen/skincare/etc, but it’s just hard to get the cadence and everything so incredibly legible. There’s a reason Scott makes the big bucks.
I notice that you toned down your disrecommendation at the top considerably in the current version compared to the email you sent out. I didn't read the post based on that.
But, I came here to comment that I generally like the book reviews, and I'm open to the idea of guest posts, but #guest_posts_are_not_a_suicide_pact; I think that if readers have things to say that aren't of high quality and considerable interest to you, then maybe you should just encourage them to have a substack of their own, and then highlight it in links or open threads?
I know this is partly a joke and I’m sure you agree with this next part based on other stuff you’ve written but for scrollers: you are one guy in one particular place and society overall thrives better if there’s some fostering of community and mentorship.
Edit as forgot to add the point: I am happy to see guest posts as I see it as pro social and and you do it the appropriate amount from my perspective.
I mean, someone who I assume is a friend of the blog or the rationalist community is going through a tough time, so Scott's boosting his writing. They don't have to be totally rational, they can be human and do something nice for a friend in trouble too.
I explicitly do enjoy the book reviews and this guest post, though I usually skip the open threads these days. To each their own. Now, if only there was a way for a RSS feed for each tag (guest post, thread, regular post) to exist and we would all be happy.
I disagree. I like the guest posts and book reviews. Especially when Scott is distracted / busy and so quality slips. If you don’t like them just don’t read them 🤷
You're missing out, and I'd recommend at least reading the past book review contest winners. The Georgism (Progress and Poverty) and Egan (Educated Mind) book reviews have stuck with me ever since I read them, and are basically introductions to a particular scholar's entire theory of economics/taxation and education respectively.
Out of this year's book reviews, I'd recommend the one on prions (The Family that Couldn't Sleep). Of the ones I read so far, it's the one that would have gotten my vote if I'd managed to actually vote in time.
You can filter out emails with a certain keyword very easily. There's also ways to do this for RSS feeds (https://siftrss.com/ is one I'm using). Just remove any post with "guest post" in it and boom, you're done! I use this for various "current thing" topics like "Gaza".
I found this really interesting and with useful insights on a problem I’ve thought about a lot in the past. Thank you for taking a chance and posting it
I agree completely with this, and correspondingly disagree with Fujimura's comment above -- I often read the book reviews (you can almost tell after the first couple of sentences whether you want to go on, which I often do!), and am alerted to things I would otherwise never have considered. Anyway, bravo to Böttger for this post; it's fantastic!
My heart goes out to Böttger - his experience with cancer was worse than mine and mine was pretty bad - but he's missing the last third of his story. He isn't "out of it yet" so he can't see his experience from the outside and judge whether the frame he put around it was a good one. As of yet, it's still a howl at the void. Recover, Daniel. Get well.
Glad to hear it! You told us your ex-father-in-law’s health status by the time of the writing of your essay, but not your own. There’s a lot of space between “fully recovered” and “capable of pressing ‘send’ on an email.”
I'm in radiotherapy, chemotherapy will come next, because that's what you do with stage 3 astrocytoma. Mental functioning is intermittent (pain-dependent), physical functioning is weak but stable. Full recovery in 2024 would be (falsely) considered a miracle.
Thank you for your responses. The fact that you’re alive to type them is, if not a miracle, then a desirable outcome of low probability. If you wished for full recovery in 2024, you’d be foolish and impatient, but you have more than just two and a half months. It’s very likely that be better in January 2025 than you are now, and even better in October 2025. At some point, hard as it may be to imagine it now, you will have recovered. You can do it.
Let's face it, you're only "out of it" when you're dead. Treatment is only a temporary relief; our bodies continue to rot until they inevitable decay into nothingness. There's no light at the end of the tunnel.
You say 'rot' like it's a bad thing. As someone who is outdoors a lot, I see rot as being a teeming - seething - universe of it's own, busy and rich with life of a different kind. Decay does not result in nothingness, but in rebirth. (Just not the rebirth of the original organism!)
Why not? It's interesting, and (as I said before) quite dynamic; there's lots to observe and think about. Is it ideal? - probably not. But it's better than the alternative, and that's got substantial value.
It looks like the metaphor of space-efficient VS time-efficient could be extended to high-bias VS high-variance models in machine learning. If you constrain a model a lot (for instance via regularization or even simply by keeping it small) it won’t overfit, meaning it will be more reliable. This can make it reliably wrong (biased) sometimes, but that’s a price you may be ready to pay given the use case. In medical use cases, the work of people like Cynthia Rudin showed that indeed a very simple and interpretable model such as a scoring system can be learned optimally and can save lives. The nurses are like that. To value complexity you need higher capacity (in a way this is thinking fast and slow all over again) and, crucially, trust.
Nurses are the example here, but this applies to bureaucracies creaking under the load of clients, resorting to form-filling because it would be bad to spend too much time on one "real" conversation while another client with a possibly more important problem is waiting.
I think you may be missing the simpler explanation: bureaucracies don’t care about people. They’re optimizing for efficiency because it can be measured on a spreadsheet and requires zero courage or willingness, on the part of the institution, to make a sacrifice.
The alternative explanations I thought of are that doctors and nurses (or any expert) will disregard any non-expert suggestions, information and opinions, for several other reasons beyond optimizing for survival:
1) Even though the suggestion might be good, it takes too much time and effort to decide whether it's good. The system runs on trust - trust that other experts know what they are doing - without having to verify each instance. This means false positives (accepting expert suggestions that are bad) and misses (rejecting non-expert suggestions that are good). This is not necessarily because of optimizing for space over time (a complicated and lengthy suggestion by another doctor is more likely to be taken seriously than a concise and simple one by a patient), just efficiency in general.
2) Experts take offense at being second-guessed by non-experts. Protecting their status takes precedence over helping the patient. They might not be fully aware of this. Patients are so low on the pecking order they aren't even on it - they are objects, not subjects or God-forbid peers.
I have seen this attitude in some of the surgeons that I've worked with, but private practice physicians with a focus on patient care over surgery that act this way are the rare exception in my experience versus being a norm.
I'm editing to specifically underline the word "some" above. Plenty of the surgeons that I have worked with exhibit more focus on the patient.
I mostly have experience with public healthcare in my country. The attitude describes 5/6 doctors I've had. I've also tried a private clinic in another country, where the doctor did listen to suggestions, but seemed to veer towards the other extreme of doing whatever it took to make me a satisfied customer rather than giving a good diagnosis and treatment.
In general it seems to me that healthcare is focused on treating the 20% of easily solvable ailments that account for 80% of the cases, and don't really do much about rare, severe or chronic diseases. Hopefully in the near future we'll look back at current times and wonder at how primitive it all was.
The thing about US care is that we're business oriented thanks to our private ins system. Docs of all stripes are incentivized to at least resemble a caring physician. Our two systems are likely more aligned in attitude at the hospital v. hospital level.
It also means that 80% of the caseload that is common problems is easily managed en masse. At least in my field, ophthalmology, that includes an array of longterm illness, especially glaucoma and dry eye syndrome, as well as more niche but still common enough concerns like high risk medications (e.g, hydroxychloroquine).
So, we refer out at the primary eye doc level when someone needs more advanced care, especially severe glaucomas and low vision services.
When doctors on duty talk to each other, they're VERY terse, in my experience. When they have a lot of information to pass along, they hand each other printouts.
I was not second-guessing them so I don't think your explanation applies to my experience.
Right, I meant more than any suggestions tend to be interpreted as second-guessing by some doctors, at least in my experience. Anyway, interesting read! I've never been through anything like this, but still felt like I could relate. Wish you all the best!
I have a brain tumour too and my neurosurgeon seemed like he just wanted to get his job done and move on but everyone else I have dealt with (oncologists, neurologists, nurses and MRI operators — even the bureaucracy of the NHS) have treated me as though I am the most important person in the world. I feel like they are all my friends and, in different circumstances, would certainly share a beer with me at the pub. They always respected my opinions and wishes.
Agreed. MS patient in Switzerland (which has a system with a highly standardized insurance mandate but some competition in care provision that in practice ends up far closer to the US than to the NHS) and while I have a lot of beef how long people did not catch it, after the diagnosis everyone has been like that.
Even the health insurance which covers a shit load of costs every year (with no obvious path aside of it increasing further) is nice and helpful!
Is *everyone* perfect? Of course not, I on one occasion had an occupational therapist which I neither liked nor thought useful in the least and that was quickly resolved (by nuking the undertaking which was exactly what I thought should happen).
Breast cancer treatment here in the US and I say the same. My surgeon was easily worn out by conversation, but every other provider -- oncologist, radiation oncologists and technicians, radiologists, and almost every single last nurse have treated me with the utmost attention and care, down to the very last detail. This at a small fairly rural non-fancy hospital with a tiny cancer care center. And despite Covid pressures and burnout and serious understaffing.
The gratitude I genuinely felt and continue to feel was an incredible existential analgesic for me all the way through. Their care and my gratitude for it felt like direct pain medication for the suffering my fear caused.
I can imagine all kinds of ways I could have produced friction for them and me. To step into that level of urgent, highly coordinated medical care is to travel to another country where you don't know the rules, don't speak the language, and your norms and ideas aren't theirs.
It helped me quite a lot to surrender to all of that and trust that they were going to carry me where I needed to go. I spoke up here and there about a few things, but otherwise the collaboration I had in mind was to participate in their system and receive its benefits. It's not the kind of collaboration I'd seek with a single other person, like a therapist or a partner or friend. I think it helped me to see what a stellar job they were all doing under such terrible circumstances. And it probably helped that I'm also a healthcare provider so I have a lot of compassion for the demands of their jobs.
I think some people do get terrible care. There are loads of people who are traumatized by their care. I was to some extent but not the fault of the providers in my case. I think some people have unrealistic expectations about what's possible in that space. And of course loads of people are just out of their minds in pain or terror and can't be expected to be in it any other kind of way.
That's simpler but wrong. Bureaucracies are the jobs of other people, and they're no exception to the fact that most people are honestly trying to do the right thing, under circumstances that you're failing to imagine.
This reminded me of reading Sam Kriss, in that I didn't understand it at all but get the feeling there must be something brilliant behind the words...probably? And there's a third tragedy in there despite only two being explicated, I think? Honestly just sort of baffled. (And worrying about my own "benign" tumor...cancer sucks.)
The various doctors I spoke to about my large tumour told me that the only way to deal with it mentally is to completely ignore it. They even strongly recommended against regular scans of it (both NHS and Private doctors so different incentives) as it would only cause worry from minor growth and any actual issue is going to be noticed first through symptoms anyway. I can't imagine the horror of finding out your benign tumour was actually malignant though.
Yeah, that's where I'm at..."well, we could schedule you with a neurologist" left hanging as an open invitation, and then subsequently worrying about random unexplained one-sided headaches in the supposed area, plus other desiderata. Maybe coincidence, maybe not...would I even notice such anomalies without the priming? Obviously I'd prefer to have a better wakeup call than crashing a car or something similarly dramatic. But all those Bayes lessons about the classic mammogram problem are very much an EY-changes-your-thinking thing. Some avoid doctors cause they're hyperchondriacs, some do it cause they don't wanna wrestle with thorny statistics problems...having had a few relatives die from various cancers isn't reassuring either, even if they were often long-lived. Probably best to just not think of the pink elephant in the brain, which isn't steerable anyway.
Is this a UK thing? I feel like in America you would get an audience with a neurologist if you had a large tumor; but I don't use the medical system so may be confused about what a neurologist is/does.
In the US it often depends to an almost ridiculous degree on who your doctor is and who your insurer is. This is true regardless of how clear or unclear it is that scans and doctor visits are helpful vs watching for symptoms.
I also have a tumour (also NHS). I researched the hell out of mine. Three hours a night for six months. I've had maybe 10 MRIs. I appreciated every snippet of information and they gave me comfort — even when the news was bad. My oncologists and nurses encouraged me in my research and answered all my questions.
Before posting the comment I tweaked some wording and accidentally appear to have removed the word benign (as I was replying to a comment about a benign tumour it mustn't have tripped my final sanity check). But yes, the NHS were very fast and helpful at getting the scans done but once it got confirmed benign they shifted into (probably rightfully) telling me to stop worrying and ignore it unless symptoms show up.
I don't know. I thought it was an interesting exploration of religious questions in a very atheistic, rationalist way, sort of a 'what if a very religious person by nature didn't believe in God'. Though as I've said above, I don't hold the same view. But that's because I'm not the same person.
Sam Kriss...well, let's say he's very good at what he does, but I find him an irritating representative of the literati class with their typical prejudices ("haha let's mock working=-class and middle-class old people because they voted for Trump"), and I'll leave it at my personal statement of disgust, which only asserts how I personally feel about something. :)
Kings New Clothes syndrome. Always reminds me of attempting to read Wittgenstein's Tractatus Logico-Philosophicus many years ago. Now we have Wikipedia and the interweb I must have another go.
Interesting that you mention Sam Kriss. Daniel's post is so honest and open. I feel that I'm reading the work of an adult. Sam's work, although brilliant, does not leave me with that impression.
Love this- thank you. As a scholar of religion (and having skin in the game in religious spheres), this tension is a fantastic predictor and pattern-matching tool. The more you value your system and feel its uniquely beautiful and transcendent, the more you have a 'survive' mindset. This is linked to orthodoxy and traditionalist approaches, which I am favorable towards. "I want to avoid this from collapse because its so great, and the way it is right now is the way it should be".
But as you value it, you also can tend towards thrive and take up a totally different approach. "But we can do this better!" "But this could help EVEN MORE PEOPLE!" "But lets update for the times, or culture, etc!" The thrivers have a similarly good intention and can come to a totally different approach/conclusion with an equally valiant wish. And its hard to empathize or understand sometimes, as it can feel 'competitive' or 'misaligned'.
So anyway- I just wanted to thank you for the essay. It helped me understand something more clearly that I have been processing for a bit and now is seemingly obvious. Wishing you the best :)
Oh, very cool. Yes I agree thrive/survive does work on the reformer/orthodox quarrels that are now ongoing in many religions because they're all suffering the huge onslaught of the Internet and sanity/atheism. There are attempts to explain this with political left/right but thrive/survive imports less baggage. I've done scholarly work and a few minor papers in the study of religion. (Main result: https://sevensecularsermons.org/why-atheists-need-ecstasy/ ) If your pursuit of thrive/survive into contemporary religious dynamics leads to a publication or something, I would love to be told.
I already had some fleeting not-quite-thoughts about this, and my gut reaction is that this is awesome. It feels like someone flipped the switch on, and I can see the hypothesis I was trying to test for. It is directly relevant for my work and I will be testing it abundantly. Thanks.
Likewise. I think about reciprocity - tit for tat - a lot, and the space-time tradeoff seems to shed light on the behavioral choices as much as the thought/reflex split. Gonna go read the Wikipedia article and then do some coding.
"My father-in-law was the front seat passenger. Same story with him: put into the CAT scanner to look for fractures, and although he never had seizures they found a brain tumor in him as well..."
This reminds me of the "hitchhiker you picked up" joke. He asked, “How do you know I’m not a serial killer?"....
This is a great concept! I love the application of computer science. I agree that survival oriented processes are fighting a doomed battle, but I don’t think you take this far enough: They’re trying to push the probability of death to zero, when any study of Eliezer knows, zero isn’t a probability.
The only solution I see is, the survival oriented processes have to trust the thriving oriented processses to run the show most of the time, except for in immediate- ie short time duration- crises. If a “crisis” or “emergency” goes on for long, it will kill thriving. Yes, each process has to trust the other, but ultimately the thriving-oriented process has to be the captain of the ship. If the survival process doesn’t trust the thriving process to ultimately keep it safe, no amount of risk-mitigation will suffice.
From that point of view you can view the early stages of human history (first hundred thousand years or two) as being more in survival mode, from which we then gradually though unevenly have been emerging in the past couple of millenia, esp the last couple centuries. The thrivers are in charge, sort of, here and there, but are still pulled down by the overwhelming flywheel effect of the survivors, who still dominate. Who knows where we'll end up, but Böttger's theory could sure help in keeping us on the trajectory of the past couple centuries.
If thrivers were actually in charge, we’d have nuclear power plants everywhere and very little regulation. We wouldn’t have masked toddlers in response to Covid, or said people have to give up bodily autonomy to ward off the small threat posed by Covid. There wouldn’t be major politicians pushing for digital censorship. We’d be traveling the stars, not fighting over nonsense.
Thrivers ARE in charge locally, here and there; they run a lot of companies and other organizations, in some rich countries, though not usually the political systems of those countries. (The current president of Argentina could be classified as a thriver.) But as soon as we get into this discussion the question of classification becomes pretty vague. Take Elon Musk, often considered a paradigmatic thriver. But when he ventured to support the Ukranians with Starlink satellites, the Chinese were pissed off and said look here, you've got a huge amount invested here and you're dependent on us, so behave, i.e. support our friend Vladimir, or we kick you out. So Musk changed his tune and started parroting Putin rhetoric, e.g. idiotic stuff about holding "referenda" in Donetzk and other occupied territories to decide whether they should join Russia. (And in American politics it's driven him all the way into his current hyperbolic rhetoric . . .) Of course -- he wants to keep what he has; he sees the democrats as a threat not only because of taxes (that's a very small part) but because he doesn't want to get kicked out of China. So is that survival or is it thriving? It's clear that this person was a high-profile thriver up to a certain point, and it seems pretty obvious that he's now turned into more of a survivor -- but one who still needs the reckless swashbuckling thriver image he'd cultivated so consistently before 2022. Hard to classify, and I think you run into the same problems of classification with many, many people. I agree with the commenter above who thought that the survivor-thriver antithesis is too one-dimensional and needs to be seen in a wider and more multimensional context.
I think the whole distinction is a bit silly. You can be as thriver as you like at 9am, but at 9:05am something can happen that puts you (temporarily or permanently) into survivor mode.
Nobody is an always-thriver. If there's a distinction to be made it's between the sometimes-thrivers and the never-thrivers.
I try hard to walk through the world smiling, friendly, laughing, generous, happy. Given a moment between stimulus and response, I can usually be that guy. Under duress, not so good. I agree the distinction is granular; not silly, though.
We had this system, generally, during the beginning of the industrial revolution. It had horrible labor conditions and massive pollution problems. It may have led to faster "thriving" in some sense, a faster advancement, but it had tremendous downsides as well. We've perhaps gone too far in the other direction at this point, but to pretend putting "thrivers" in charge would lead to some harmonious existence is not supported by history. Many people would get stomped on underfoot to advance their goals.
The last 200 years has seen a monotonic increase in American life expectancy (possibly excepting wars, which some American data may not be granular enough to show.) While England essentially forced its rural populations into cities, in America people went voluntarily into cities and factories. Every indication is that, at least for the span of the early Industrial Revolution, the benefits of rapid industrialization far outweighed the costs. Pollution was just not as bad as starvation, material deprivation, or losing the war. And farms of the Industrial Revolution, frankly, had terrible labor conditions as well.
That we can have a cleaner environment despite our oversized population is a direct benefit of our developed industrial capacity. If modern Americans lived 'off the land' in a 1700s manner we would quickly denude the landscape beyond the point of repair.
Perhaps we're at the point where marginal improvements are not as valuable so we can afford to reign in the investors. Or perhaps not. But during the industrial revolution improvement in the economic sense was absolutely a strong net positive, with long term residual benefits. I'm not sure if the investors should be credited as strivers or survivalists, granted. But whatever happened during that period improved society in the short and long term.
It might be too late if you wait for the emergency to arrive. For example, many European countries switched to full thrive mode just before the rise of Nazism, and so did not have time to prepare themselves when attacked.
Optimally you would need to keep some places and some people on constant survival mode, so there will be some help ready for an emergency.
There's a venerable German state theorist whose name escapes me, who says the frue sovereign is the one who can declare a state of emergency. Maybe that has to be a survival-oriented person/institution.
I see it exactly reversed: Most processes value thriving at the expense of survival to a suicidal degree.
People will burn every tree on the island to smelt copper, then freeze to death in the winter 100% of the time unless the few survival oriented perspectives fight a brutal battle of the somme trench crawl against the optimists forever, every day, because there is a limit to how good something can get on a set time scale but there is no limit to how bad something can get. Goodness/badness is a multiplier on a variable, and the multiplier for death is 0.
I think the OP dichotomy is between short-term survival and long[er]-term thriving. The issue as you point out that both the post and most of the society do not pay enough attention to long-term survival.
I was thinking more metaphorically, but thanks for the link. That bit always struck me as odd: It's easy to imagine overhunting to extinction, but you can get on a high place and use your eyes to see how close you are to tree = 0.
Surely there was at least one survival maxer to throw a fit around.
A rivetting piece! The perennial conflict between Surviving and Thriving is both new to me and intuitively correct, and also something to incorporate in my worldbuilding.
However, Urgent communications are always terse, whether Surviving ("Hold the line!") or Surviving ("Charge the flank!"). So the first rule has to be: *Recognise when the situation is Urgent to the other parties.* That's not easy when medical professionals are projecting calm, and maybe impossible when you're doped on meds and distracted by pain.
I'm intrigued by Daniel's observation that Surviving conversations are terse even when Non-Urgent.
Rather than "space efficiency", I think this is about Survival being a complex but solved problem (except when it's suddenly not) - basically Chesterton's Fence.
If so, then people responsible for Survival are terse because the conversation is too big for the time and energy available (you can almost hear the mental sigh just thinking about it), and because they feel an instinctive need to maintain authority.
Spatial efficiency usually comes at the cost of temporal inefficiency. And there’s no way I’d describe survival oriented persons as being “patient, willing to spend long amounts of time”.
I think the (unstated) assumption is that the time-inefficient component of survival happens before the crisis starts. Then, in crisis, it relies on short, cached phrases.
In other words, doctors and nurses have years of medical training so they can communicate efficiently in the moment using standardized language.
Yes. So if you said to them, "Hey why don't you pack an extra beach towel?", their first response would be terse and unenthusiastic. They can't even remember *why* they've packed what they've packed, but they know it's optimum and really don't want to revisit it, and if they do it will be a whole process to decide to make room for the towel and then make that work.
Hmmm. I like the idea of understanding the difference between the survive and thrive mindsets, and the idea of understanding where others are coming from and being charitable to them as a way of fostering better cooperation. I'm not quite sure I'm convinced of the difference between survive and thrive mindsets mapping easily onto the difference between space-efficient and time-efficient algorithms, though. From my reading that seems pretty specific to your experience at the hospital, and really only one specific part of the hospital experience (their communication).
I feel like there are a lot of other things that go into the difference between the two mindsets (e.g. level of risk aversion, or in this case, different incentives: the hospital mostly doesn't want you to die, but cares very little about whether you're having a good time.) that to me cannot be explained away as a difference of algorithms. I don't think that e/acc's issue with AI doomers is that doomers are communicating too curtly, and I don't see how the third heuristic will help the issue. I do, on the other hand, see how the second and fourth heuristics could be valuable, but those seem more about just being charitable than having anything to do with running different algorithms. Maybe I'm missing something here, though, and would be appreciative if anyone can explain what it is.
I think that e/acc's issue with AI doomers is EXACTLY that doomers are communicating too curtly. "Orthogonality", "instrumental convergence on power-grabbing" etc. get used as if they meant anything to people who have failed to Read the Sequences.
But they're all IT people, they understand Algorithms 101, this should help them get to the bottom of their failures to collaborate.
I'm not sure you've convinced me. I can see why using jargon/abbreviations/terseness would be frustrating for that, but I don't think translating that jargon into a message that's more time-efficient and less space-efficient would make any more than just a marginal difference. And it's not clear to me that survive mindset groups are any more likely to communicate tersely than thrive mindset groups (I've seen a lot of very very long blog posts from AI doomers), with the notable exception of a situation like a hospital where things are happening very urgently on a timescale of minutes or even seconds.
It still feels to me like the main thing separating these two groups is risk-aversion, priors about how likely AI is to kill us all, priors about how great AI will be if it goes well, etc. (This is my attempt to be charitable. There is part of me that thinks a lot of (but not all) e/accs are just trolls or people who don't understand tail risk or survivorship bias very well.)
Yes. I don't use "collaboration" and "communication" synonymously. Reading each other's long blog post isn't usually collaboration. Collaboration is when you try to solve a problem together, and that's where the communication styles diverge.
Yeah I found no problem with the survive vs thrive dichotomy, but half way through I thought the answer is short-term vs long-term threat aversion, which kinda maps to his "space vs time optimization", but not really. I think my answer generalizes much better though, lots of problems stems from not knowing the future, and trying to handle that.
Thinking about this more: I don’t believe survival processes could be optimizing for spatial efficiency at the cost of temporal efficiency. There’s no way “take longer to do the processing” mitigates the risk of “bit flipping” from taking up more space, because in an emergency time is scarce, and reaction time can be life of death.
I think we should expect survival oriented processes to just be computationally simpler period. We should should expect them to be error prone to the side of over-estimating threats. Seeing a threat where it doesn’t exist inhibits thriving, but not seeing a threat could kill you.
The personal interaction rules you give are a good description of “how to deal with an emotional person”, as well as “how to interact with a computational process with very limited capacity”: send short messages, as few as possible, and be patient.
I think you underestimate greediness of such a choice. Even if _total_ time is shorter, each individual _step_ of a fancy algorithm is often longer and/or more complex, hence more prone to disruption. (And yes, what counts as an individual step is fractal, yadda-yadda-yadda.) MergeSort's individual steps are efficient but more complex than "compare A to B".
The prone-to-overestimating part is, of course, true.
I symphatize with his attempt to crystallize what must have been a phenomenologically intense experience, pain notwithstanding, like, it's *that type of experience* that's notoriously difficult to communicate.
Given that certain substances like LSD or ayahuasca can scramble our sense of salience (i.e. what feels important or significant), I usually take shifts in perceived importance with a grain of salt, especially from people who have a drastically different epistemic standard. But some of the readers here, including me, would be predisposed to extend extra trust to Böttger, given his background and past work.
For what it's worth, IMO the post managed to communicate a very real and very important thing, in particular the seven heuristics that probably will get quoted a lot in the future. It needs more elaboration/response by other people (or by himself) in order to gain a stable position in people's metaethics, but as far as memetics go, it could spread far and wide.
Some quotes that i like:
"Whenever these people think there is a minute for idle chat, that proceeds flawlessly. But the more urgent collaboration is, the more frequently it appears to fail"
Reminds me of the distinction between no-slack work vs. work that has slack built-in e.g. film industry.
"The brevity of their communication will feel hostile; interpret it charitably as an expression of urgency of concern."
This resonated a lot, although gotta say: My experience living in a non-western country, where scarcity is more common, is that brevity isn't just about urgency and concern but also reflects a fear of losing out, so it's also a kind of a power play—though that might sound cynical.
A very interesting point, and one I will try to use when analyzing situations in the future. A couple of other places where it seems relevant:
Teaching. The teacher is on survive mode, knowing how little time is left to cover the entire material, the students want an explanation of why they have to learn it for every single new algorithm.
Relationships, when one partner have more experiences with relationships breaking / divorces, and so try to optimize for survival and the other partner is more afraid of the relationship going stale and so optimizes for thriving.
Probably will think of others soon. Also I for one really enjoy having guest posts here, I usually find new things to read mainly through links from blogs I already read. Will try to look into your sermons when I have time (also, a movie recommendation: https://www.imdb.com/title/tt0418455/ )
Having a brain tumor is a really horrible stroke of luck, and I'm really sorry for you. I'm hoping for a miracle.
Ironically, I get more of an Epicurean conclusion from this.
What this makes me think is that I'm glad I never really spent that much time on philosophy, and I feel better about all the time I spent trying to get laid (and...adjacent activities). You engage in all these intricate mental computations and constructions, ponder the meaning of life, and it turns out to be just a tumor.
(God, that SUCKS.)
Of course people enjoyed them and found them meaningful, so I guess that's something. But to me it further lowers my estimation of the probability of the Divine, or any kind of transcendent meaning to anything. That feeling you get? Yeah, it's just some brain circuitry firing a certain way or, as in this awful case, a tumor.
I just never *got* religion. I had zero interest in Zen or ego death or anything mystical. I understand the desire to save your soul from an eternity of torment in the afterlife, but if that isn't true...what's the point? It seems like it serves other people more than you. I think I just don't have whatever the temporal lobe wiring is for it. Or something.
Eat, sleep, work to pay your bills, f***k if you can (unless you're ace), save for retirement, raise kids as so many pronatalists want to so you have something left after you. (I whiffed on the last one, though I admit I was never really all that interested and shocked people in high school by admitting it to them.) The rest is commentary.
You're just a monkey with a bigger cortex. I am. We all are. Enjoy your bananas, the end is coming for you sooner or later.
But: having a brain tumor is a really horrible stroke of luck, and I'm really sorry for you. I'm hoping for a miracle. I'd pray for you, but if He's up there, I doubt He'd listen to me.
I came to the opposite conclusion: I have to meditate enough so that I can withstand the tsunami of suffering which life will at some point dump on me. I didn't focus on the shortness of life, but on the long suffering at the end.
Also, maybe I shouldn't have kids, so I always have an exit hatch. Although, to be honest, I would not blame my dad at all for taking his life under these circumstances (not sure of the exact details of OP's situation, sounds like the mother is no longer around which does make it worse).
Of course whatever works for someone right? We don't all need to be walking the same path.
Do you know My Stroke of Insight by Jill Bolte Taylor? A neuroscientist reflecting on her altered perceptions from a massive stroke and the insights it led her to that were life-changing for her.
To me a brain tumor or a stroke or psychedelics providing a dramatic shift of perspective doesn't undermine the credibility of the perspective. It says to me that the normal well defined circuits of our ordinary thinking lead to X kinds of awareness/knowledge and exceptional experiences can produce extremely interesting Y kinds of awareness/knowledge. Like the difference between research studies and poetry. Are they not both valuable?
He isn't up there. (Yet. God growth mindset!) Save your hope for likelier things.
I do continue to think that in a sense, we're also all the same universe that's wearing our faces as its masks, using many brains and networking them together through language to figure itself out. That's just philosophy, didn't go away when the seizures went away.
People often ask how the methods of rationality can help in our daily life. The first part of this essay was an eye-opening insight into how rationality can help, when it is all you have left.
Very much so. My fellow patients were busy with painful thoughts like "this is because I'm so stressed" or "I knew it would all go wrong if I did not quit my job" at exactly the worst possible time for such shit. The Methods of Rationality seemed indispensable for navigating this.
Like some commenters here, upon reading the point about the survival/thriving tension, it intuitively felt correct to me. It is a simple but insightful theory. Thanks for sharing, and I’m sorry you had to go through such pain to discover it. I am amazed you were able to break through the pain and share it.
The theory sheds some light on a tension I’m experiencing between two groups in my life. I am part of a local YIMBY movement and have had several long conversations with local NIMBYs. Often time it feels like we are speaking different languages. I could not understand some NIMBY concerns, like how tall buildings would decrease the sun on their gardens + simply not liking tall buildings, would be in anyway comparable to the huge exodus of people (many of them my friends) from my city because of the lack of housing.
Instead of just concluding these NIMBYs are garbage, Böttger’s theory gives a much kinder, much more actionable explanation for what’s going on. The YIMBYs are motivated by survival (i.e. people able to live in my city) whereas the NIMBYs have their housing and have now progressed to focusing on thriving.
Pattern Matching!
I don’t think the time-space part of theory is correct, though. In my example the tension seems to be between (YIMBYs) the existence of housing and how to get to a minimum where people can even think of living in my city vs (NIMBYs) improving the quality of housing for those who have it. The idea of space efficiency doesn’t really fit in with ensuring increased housing supply, at least not to me.
What makes more sense to me is not a space-efficiency tradeoff, but rather a time vs. solution-quality tradeoff. In more computer sciency terms, it’s the tradeoff between the faster but less optimal approach of approximation algorithms vs. the slower but provably-optimal approach of the algorithms we learned in Algorithms 101. I’ve seen this dynamic most clearly in my upper computer science classes on NP-Complete approximation algorithms, which you can read about here: https://www.khoury.northeastern.edu/home/rraj/Courses/7880/F09/Lectures/ApproxAlgs.pdf.
Finally, I think Böttger’s suggestions at the end still apply regardless of which theory proposed by people you choose here. I would just add one imperative to the Team Survival’s checklist, which is to try to convince Team Thriving that this is indeed a survival situation, because survival is more important than thriving. If the situation is indeed a survival situation, the Team Thriving should be able to temporarily set aside their goals, and join forces with Team Survival. Maybe we call the synergy of the two Team Flourishing?
The YIMBY and NIMBY labeling is surprising! My first intuition is (YIMBYs) want to optimize the current system to improve the lives of mostly strangers that can be doing much better vs (NIMBYs) want to keep everything exactly the way it is, we are doing fine and doing anything different might hurt, thank you very much.
I doubt it. Despite the label I think that "YIMBYs" are rarely people who actually own a backyard, they're either people who are desperate to move into the neighbourhood (and don't care if it gets slightly worse in the process) or property developers (who probably live in a different place anyway).
You could also reverse the survive-thrive classification here just as easily. NIMBYs are "survive" types, they just want to maintain what they already have. YIMBYs are "thrive" types who are trying to move up in the world, either from tenant to owner-occupier or from resident of a less popular place to resident of a more popular place.
Alternately, NIMBYs who worry that the wrong sort of people moving in and ruining the neighborhood could be said to have a "survive" mentality, whereas people who expect a neighborhood to improve with more development have a "thrive" mentality.
Id versus Superego. Some people are more driven by one, some by the other. It maps onto this theory and explains phenomena like the masochistic guilt shown by some groups to others (overpowered oppressive superego they identify with) and the sadistic glee by which those others concur with that guilt (overpowered wild Id with rejected superego.) There are many other terminologies for this core individual human conflict and how it might be projected out in various ways, but currently I find this one most elegant.
Is this not just short term vs long term goals, tactics vs strategy? Survive is tactics, thrive is strategy.
Every specialism has its jargon, primarily to avoid long-winded descriptions. This appears terse to the outsider but it's just efficiency. Knowing the lingo creates an in-crowd, especially if they're working together every day. Plus a medic needs to distance themselves to remain objective and do their job. These can combine to make the patient feel excluded. When you speak it's as if you're a distraction from their private business and it makes you feel like a child interrupting a grown-up conversation.
[I realised after posting this that doctor-patient is classic parent-child in TA]
re sorting books: surely you just repeatedly start from the beginning, swapping adjacent books if they're in the wrong order? ;)
I know too little about military things, haven't even read Clausewitz. Do they have a standard resolution mechanism for differences of opinion between tactics and strategy? I can imagine armies where strategy always wins via chain of command, but I can also imagine armies where the folks in the trenches decide what elements of the strategy are actually doable. War gives pretty intense feedback though, so over the millennial, centuries and decades I would expect some kind of convergence.
In military terms (and I know less than you do) I presume strategy sets your objectives and tactics is how you achieve them. But "No Plan Survives First Contact With the Enemy".
re BubbleSort: if you were in a wheelchair and all your library was on a single floor, that'd probably be the most efficient. My tangential point was that algorithmic efficiency isn't fixed in stone: it's about having a large library of possibilities to choose from.
Having read the theory, I feel like the author's belief that it will contribute to saving the world is a grandiose delusion. In light of the line "Although I can’t speak for its world-historical importance", I infer that Scott also thinks this, although he is surely too polite to express his doubt in those particular words. This causes me to suspect that Scott has published the essay primarily out of sympathy for a very ill man who describes himself as adjacent to suicide (no one says "I can’t do that to the kids" unless they've thought about it real seriously).
I know it's very cruel to say, but I totally agree with this. To be fair to the OP, maybe he meant "contribute to saving the world" in the same way that donating $100 for mosquito nets contributes to saving the world--it's a small contribution, but it is a contribution nonetheless. That said, I don't believe this post contributes as much to saving the world as donating $100 to charity.
I'm torn on whether I agree with this. I share your view that the idea that this is world saving is delusion. I'd imagine some heightened sense of salience helps with the pattern matching but also tends to over-assign a sense of importance to those patterns once identified.
In principle, I'm against pity-posting, but I do think this is an excellent essay. I found it an interesting theory written about in a clear and engaging way. I just don't think it's a world saving theory, or anything close.
If his premises were correct, perhaps it would. The delusion, alas, is a common one: that this 'thriving' as described exists or could exist. It doesn't matter if one carries this delusion forward into 'fully automated luxury space communism', the error infects everything.
'thrivers' aren't mistaken that we 'survivors' are hostile. It's not a communication problem, but a reality problem.
Whether you agree that the piece is interesting or useful or not, it seems unnecessarily harsh to say Scott published it out of pity. The OP is a decent writer telling a powerful story and offering up a theory he took from it. It doesn't need to save the world to be an interesting offering. Way less interesting things have been posted in this space.
This is exactly why I wrote the first part of the post. You should doubt this theory, because I was definitely not thinking straight. That's why I repeatedly emphasized that in the text.
Still, at the same time, it is also an idea that might make sense, might be trivial even, and maybe the only reason at has not been described before is that nobody had the same weird combination of circumstances (intensive care, Algorithms 101, nurse mother, having read Scott's thrive/survive, compulsive theorizing etc.) and time to write about it.
It was not pity. We've exchanged businesslike emails over the previous guest post. I've met Scott for all of five minutes, years ago, and he said he "kind of hated" the theory I had back then.
If I were striving for accuracy in describing delusions, I would say that "delusional" is a property that applies to people's thought processes rather than belief content per se. And delusional thought processes exist on a continuum (or share the space of, I don't want to imply unidimensionality) with non-delusional insight-generating processes. Delusional trains of thought may pass by reasonable and interesting stations before hopping the tracks. Ideas that have been kicking around your head start getting integrated, and sometimes you hit on novel solutions to actual problems. So I don't think you can exactly prohibit the publication of delusions, and I don't think you should prohibit the publication of those who are delusional.
But I agree with others that the author's assessment of the importance of this piece (I hesitate to say "idea" because I'm having troubling isolating one core insight--sorry Daniel, I think you're throwing off lots of sparks and heat but I'm concerned that it's secondary to the collapsing structure of your conceptual space) is way off base, and some features of the writing are very familiar to me as an occasional rider of runaway trains. The ideas expressed are cool tools but not paradigm shifters, and the way they are strung together in this piece makes me worry for the author (sorry Daniel, it's not pity, just worry, and I'm glad you have good German doctors and all but you're exactly the kind of person who would excel at concealing a burgeoning psychosis). But I can think of good reasons to publish in spite of this, and by my lights it doesn't take much...affirmative action for the sanity-challenged to bring it up to par, if that was even a consideration. I wonder if Scott's motivation wasn't in part to help them both (thinking of the parable of Sally the psychiatrist here: https://slatestarcodex.com/2017/03/24/guided-by-the-beauty-of-our-weapons/) test their expectations against a larger group of people.
"Saving the world" is a grand and high-variance claim. If you're not completely on board with such a claim, but still think there is value in the perspective, it makes sense to hedge againt the high variance part. It looks like Scott did that with the intro.
I found this essay extremely interesting and insightful. Powerful even. I read this one all the way through; it was gripping. I sent it to family. Will it contribute to saving the world? Maybe! Most things don't, so the prior is low, but many big ideas that eventually work start out as low probability ideas.
This... seems very explanatory. A simple but powerful theory, with enough secondary correlations ("the Moore Law will just keep working, so we can be cavalier with memory") to be convincing. And very ACX-y.
This is an alternative framing of the old problem of decision making given certain risks and certain opportunities. Be overly cautious and you'll be "survival" biased. Be focused on chasing the shiny opportunities and you'll be "thriving" biased.
It's clear we don't frame our policies, societies and behaviours following this dichotomy, but just determine case by case what to do - especially considering that assessing risks and opportunities is far from an objective endeavor.
So I don't think this framing is particularly interesting or useful.
If you think this is just a framing, you don't understand Algorithms 101. This is an objectively solvable math problem. In fact much of it is already solved. I said there are methods for integrating algorithms of these different types and I meant it.
If you want to use this, don't focus on the meta-levels of societies and policies, go to the interfaces of collaboration and do math.
I happen to be a math guy. You got me curoius: what kind of math problem is this - except for the one I already saw: a game theoretical, utility-optimization, subjective one?
PS: framing here doesn't have any specific meaning as you seem to imply. Framing is the "kind of representation" we give to an issue.
I have to say, with all the talk about tumors and communication, my main takeaway is that a vehicle with five occupants hit a tree at 60 mph, and EVERYONE walked away. There can't be a better advertisement for whatever vehicle that was.
That's an advertisement for the Survive mentality that drives vehicle testing regulations. If the regulations had been the product of Thrive mentality he'd have hit the tree at 120 and nobody would have survived.
If we had an actual survive mentality, we would ban cars because of the absurd amount of casualties it causes. What we have now is an optimum. The increased efficiency more than justifies the deaths it brings.
Obviously it doesn't matter what dead people think, but living people are far more useful than dead people. That still doesn't mean the optimal number of preventable deaths is zero.
"dead people" who didn't actually die would be living people so your utility argument doesn't wash. What I'm saying is that there's nobody to present their point of view, not first hand anyway (and the rest is hearsay).
The same happens with war: you only ever meet the survivors so you inevitably get a one-sided view, and even that is highly selected for optimism.
Honda Civic. Yes belts and child safety seats and airbags are awesome. First responders too, and the Good Samaritan who called them and pulled me out. It's not all just the car.
I am preregistering my attempt to solve a personal, unsolved-for-twelve-years problem using the thankful theory. Will report results. If no results are reported for 7 days, I probably forgot to do it and the lack of a follow-up should not be taken as evidence for or against the theory.
Me and the other party agreed that the theory describes us surprisingly accurately, but results would probably take months of better communication to become noticeable. We are optimistic though! Might report later if I have anything unusual to say.
I'm not a doctor but, I'm not sure there's a good way to do this. If you don't have any symptoms (or some other risk factor, like age/family history), a doctor probably won't prescribe a scan either for specific body parts/diseases, let alone your whole body. As far as I know, such scans would show lots of random, weird-looking anomalies, most of which mean nothing at all. The evidence that you have cancer would be very weak compared to the prior that you don't, unless you happen to already have something that's progressed a bunch. So doing this scan would be pointless, and possibly even counter-productive (i.e. you might end up trying to get dangerous surgery to remove a harmless cyst).
(I could be wrong about the above, but it's the impression I have, )
I'm pretty sure some Youtuber went into great details about the horrifying consequences of testing for too many things, and then trying to treat problems found, but which aren't really a problem.
But this was not the video I am thinking of. The video I'm thinking of he mentions a women who has a tumour in her breast, and gets a mastectomy (or something) and has complications, and it's awful, and she really should not have bothered.
Basically... when you get tested, even if the test is mostly inconclusive, if there's something that COULD be bad, no doctor will recommend against further testing and procedures, since that might lead to their being removed from the field, whereas suggesting extra (in the end harmful and/or useless) testing and procedures will never result in the same loss of status or job.
So. Bottom line: don't get tested unless there are symptoms or family history.
Thanks for the link. The statistics are something I already knew, but it's good to know that at least one doctor thinks about this question in the way I would expect.
Doctor Vladimir Alipov (Dysphorra YouTube channel) described the same on his streams. Your description is very similar to what o remember from his streams.
I also have a glioma in my temporal lobe and I also have the same history of many (10? 20?) years of ignoring symptoms and having doctors send me away saying nothing to worry about. My tumour is a rare form called gliomatosis cerebri which grows to multiple lobes. I am up to six lobes and two hemispheres now.
I was officially diagnosed 2½ years ago. My neurosurgeon wanted to do a craniotomy but I declined because of the risk of probable defects and the certainty that he would not get all of the tumour. Up until this year, my symptoms were only minor and annoying but I have recently started to have massive seizures and memory problems. I am lucky that I have had no pain.
I'm not sure I frame my experience in terms of thriving and surviving. I decided very early on that survival was not important to me. I needed to survive long enough to make sure my wife and family are ready to manage without me. In all this time, out of a couple of hundred friends and acquaintances, I have only had two who insisted that I should be focused on trying to survive. The rest accepted and respected my decision.
Thriving has not been at the top of my priorities either. I am starting to lose my memory (which sucks) and I know that I have loss of all my faculties ahead of me but I am OK with all that too. My wife likes to pretend that everything will be OK but — even there — she respects my wishes and communication has not been a problem.
FWIW I am also a software engineer and I don't think of my experience in terms of time- versus space- efficiency. It never occurred to me and, even now, I don't think it applies. I am not optimising for anything. I'm living my life the best that I can and one day it will end. I am good with that.
Back in the paleocomputer age it was a daily consideration. Save a few bytes at the expense of more instructions in rarely executed code, spend a few bytes to save a few instructions in frequently executed code. That was many systems programmer's daily life.
I think you are right! I do this in my life and in the software I build too.
I always want to build the "best" system that I can. Time-efficiency or space-efficiency only rise to the top when they become concerns. It's usually a tiny percent of code that causes efficiency problems and we can fix them when they become a problem.
I was also there near the end of the paleocomputer age. I think a lot system programmers bring along those paleo- ideas when they are not necessary.
"I think a lot system programmers bring along those paleo- ideas when they are not necessary."
I'm sure you're right. There's personal pride in writing tight code.
The more important target is maintainability and future proofing (and nowadays freedom from malware). So many programmers just grab the nearest module off the web to save "reinventing the wheel" and then two years later when it disappears eveything goes belly up or you find all your data has appeared on the dark web. Long live NIH! ;)
I strive for maintainability and user-friendly rather than fast or small. Of course, I like my code to follow the basic rules of efficiency but optimisation can be put off until it is needed. It usually isn't.
I started my IT studies in 1998. There were many more computational resources than before, but the institutional/professorial memory of needing to mind your memory and execution time was still there.
There are still many areas today where these considerations matter, for professionals and hobbyists alike. Computer sciences. Data engineering. Video games. Server software. Trading software. Granted, time efficiency is usually more important than space-efficiency nowadays, but there are examples where the latter still matters, such as the demoscene.
Hadn't come across "demoscene". I see it's a thing (and not just the name of some forgotten geological epoch! Demos means people so it would be an alternative to anthropocene)
It's the demo-scene. A scene of people who show each other impressive technical demos, especially developed under tight contraints, such as less than xx bytes of code, or running on very restrictive legacy hardware.
Unless you are dealing with huge datasets like Google or Amazon, writing basic, efficient code is sufficient for most applications. You can optimise it later if it needs it. It usually doesn't.
As an algorithms person, your space vs time analogy was very muddled. In a basic model of computation, algorithms aren't at risk of having their memory overwritten. There's no fundamental reason why robustness should equal space efficiency.
I have a similar objection to the survival / thriving analogy. Yes, there are some parallels that can be ported over, but it doesn't really match the immediate vs intellectual modes of problem solving or communication.
That said, your model does make sense. I know viscerally the clusterfuck it can be when trying to wrangle with urgent / high-stakes situations using intellectual communication norms. Effective people need to learn to switch between these modes as the situation calls for.
> There's no fundamental reason why robustness should equal space efficiency.
I disagree. Random damage - minefields, artillery bombardments, suppression / saturation fire from a machine gun, cosmic rays corrupting a storage medium one bit at a time - can be expressed in terms of average spacing between hits, that is, how large a contiguous uncorrupted area can reasonably be expected to remain.
If a soldier lying prone in the tall grass has two square feet of cross-sectional area in which he'd prefer not to be shot, and that whole field has been hit with one bullet or fragment per square foot, average soldier in said field will have been hit twice. If there were a large number of soldiers, some may have survived (probably about as many as were hit four or more times), but not enough for the overall group to remain effective. Quadcopter drones with less than one square foot of vitals each might see a far higher survival percentage, thus more overall functionality remaining after the shooting stops, under those same conditions.
Similarly, if a given algorithm needs two megabytes of RAM per iteration (that is, before it has enough progress for some sort of quick parity check to provide a useful answer about whether it succeeded, or failed and needs to be redone), and the RAM it's trying to use has an average of one unrecognized corrupt sector per megabyte, that algorithm will probably be disrupted far too often to accomplish anything. Something more space-efficient, capable of making provable progress within less than a megabyte, could still function under those conditions.
Digital systems yes, but I doubt brains are that sensitive to a few faulty neurons (or LLMs to a few faulty links) tho I guess it might come down to how the fault manifests itself. The brain must have ways to deal with neurons that get stuck in firing mode - I guess the recipient neurons just get desensitised.
I've had a couple crises of my own (different, and probably less excruciating than Daniel's), and this space- vs. time-efficiency model fits some of what I've observed in them.
In the future, I'll have to explicitly optimize for space-optimized strategies when in them. That might help quite a bit; the feeling of my empty bookshelves shrinking randomly when pain hits is familiar.
I think this is a spectacular post. I need to think about whether space-efficient vs. time-efficient is a good dichotomy, but it's at least promising.
For what it's worth, I didn't find the set-up to be too slow, and I didn't anticipate the conclusion. I don't have a background in computer algorithms.
I am glad to have read this post. It ties in very well with the Principle of Charity that SSC was originally based around. I don’t think it was necessarily groundbreaking, but perhaps that is because I was already practicing its lessons but others were unaware.
This is a "galactic algorithm"(pejorative term) concept. You can find good programmers who dismiss the concept entirely and I bet you could dismiss it as highly improbable that the brain is optimized on the space frontier, given that good compression seems to be serial and very fragile, your evidence is the brain recovering data and the brain seem fractal and highly parallel.
> Space-efficient communication can’t cache a long message to be communicated, or a long message received to be understood. Therefore space-efficient communication has to rely on short, atomic messages, which in order to be informative have to be pre-agreed.
Im pretty sure this is just flatly false. The best compression uses sub-bit entropy, any missed bit causes a cascading misalignment of plausible data.
> the wetware it implements usually one, sometimes two, phonological loops
My reading of such things is that *every* neuron is a loop, and that the structure of the brain is very loopy, I dont know where your getting the claim theres 2 loops from given theres at least feedback loops from chemical systems; to be *extremely crude* your sleep cycles depend on a daily cycle of hormones that float around the brain, and idk what the mouthy cycle for women is based on, but it seems to be ... monthly and a cycle.
It take me hours to refind the sources but I vaguely remmeber someone attempting to simulate 1 neuron with an nn and finding it takes 34 neroens in a feed back loop to replicate the behavior with some degree of accuracy; and there was someone mapping out a monkeys visual cortex as if it was an electrical diagram and it was a giant mess.
> Don't think compression.
Why? Its the space efficiency frontier; would wouldnt the data be compressed.
If you disagree with my first argument, how about a 2nd, its unlikely that brains are space efficient because compression is energy inefficient, and running on physics should provide cheap shortcuts of computational value. Consider ant pheromones and ant path finding, they use physics to outsource decentralized path finding to a massively parallel diffusion "matrix"/air that does allot of free computation for the ant, but not for our machines. Why shouldn't brains use hormones as lazy, slow, probabilistic computation systems, much like ants for with the air?
Yes the brain does lots of loopy things, that was literally my previous guest post. But not every neuron is a loop, only a few "pacemaker neurons" that do essential stuff like the hardbeat. The phonological loop is different tthough. It takes a lot of neurons, leaving little for another. You can hear a podcast while singing a song that you know by heart, but it's effortful.
Every cell has a complex electrical interaction on its surface and if you seen Michael levens with work, this bioelectricity is responsible for communicating body shapes and it's breakdown is a cause of cancer. If you cause a large splash on a small enough sphere doesn't the wave oscillate back and forth until consumed by friction?
Much like the ant, not all this computation will be useful(the air pockets where ants are bored of are still defusing pheromones) but it's there and science has not yet recreated a human brain in a lab so declaring elements the computation as irrelevant is preemptive.
I get the feeling that the author lacks insight, somehow, into the kind of situation a hospital is. Maybe because of the tumour.
I was in the emergency room myself recently (thyroid condition -> rapid heart rate). I guess British hospitals can’t be that different from German ones. Some observations…
1. Everyone is being nice to you, the patient. They are, however, allowed to yell at each other. E.g. I am given to understand that each bay in Resusitation ought to have, among other equipment, an sp02 monitor. I strongly suspect that out of my earshot, someone, somewhere is being yelled at by a nurse because the expected equipment was not there when she wanted it. Sp02 monitor is rapidly obtained…
2. They’re really trying to move patients through fast, and at least some of them are in some severe degree of danger. Like, there is at least me,, with a heart rate that is starting to look really concerning … some guy who had a seizure while driving a truck … guy who is having a more than averagely bad time with covid19 … plus, like about sixty other people who are being sorted by the triage nurse. Game is to get through the list before they die, and you, the patient can help a bit by going with the flow. I am kind of impressed by the efficiency, to be honest.
Me too about being impressed by the efficiency in my hospital experiences, and that the OP's idea of efficiency seems to be on a different track. I can't quite tell but it does seem like he was driving himself (and perhaps others) mad with his frustration about/need to understand why things are organized the way they are. Meanwhile, almost all hospitals are having big staffing shortages and staff have way too much to do in any given minute.
A large system can't optimize for everything and so at the individual point of care, the patient is going to experience things that seem amazing and things that seem kind of nuts. Some of the things that seem amazing aren't and some of the things that seem nuts are. But the main thing is that the individual patient is stepping into a vast and fast-moving system of huge complexity and they are for that time a visitor to an alien planet, albeit one that is aimed at saving/curing/helping them.
If I was to say one thing about the UK hospital system that seems kind of lacking, it’s that the electronic copy of your medical record and MyChart are not as synchronised as you might hope. In the annoying situation that the person treating you can see your medical record on their computer screen (but not your MyChart record) and the information they want is in MyChart but not your medical record, you can of course use your smartphone to show them the relevant bit of your MyChart entry…
I haven't had this experience of the distinction between my medical record and what's in MyChart, though often my bloodwork results come through to me faster than to the doctor and so we review it on my phone. Maybe that's what you mean?
One thing I think they need to work on is that all lab results get reported in MyChart separately from the doctors seeing it and that means devastating and confusing and terrifying diagnostics can come through and you may wait hours or days to talk to a doctor about it. That happened to me and it was pretty awful. They put a warning up top about that, but it seems to me they need to solve the problem at the root by requiring the provider to put a note on the lab results or place a call to the patient before the patient is staring at the results.
In the UK, the NHS “spine” has an electronic copy of your medical records that most people treating you will have access to. We *also* have MyChart, which does not necessarily contain the same information…
I’ve been living in Germany for a couple of years, during which I and a partner have both had operations (non-cancer-related) and I’ve dealt with many established systems with ways of working. I also work in global biopharma marketing strategy and have observed a few dozen interviews with German oncologists. I’m not a deep expert in German culture or German medical culture. But I can suggest with moderate confidence that German physician-patient communication in a serious-disease context is meaningfully different than in anglo culture, in the direction of less focus on listening to patients and ensuring they feel heard, and of a less patient-centric mindset in general.
That is not to say German HCPs are callous and ignore their patients. I encountered compassionate nurses, and one overnight attending who actually solved an interesting problem (pain not from core condition/op/mgmt but due to a cervical disc issue). But it’s extremely plausible that front-line providers in the ICU are, not just terse, but straight/up brusque, even rude, to patients. ICU is a situation where the system is already painstakingly and rigorously optimized and decisively outranks any attempt to fine-tune or tire-kick locally, especially by a layperson, as a matter of common sense, and that “common sense” is very powerful in Germany while the patient-centric imperative that might partly counteract it in an American hospital is weak to absent. Even in non-emergency settings, like outpatient visits during a course of chemo, the patient generally does not have much of a collaborative role in treatment decisions Germany.
I like the idea of survival mode as prioritizing space efficiency in the phonological loop - in other words, when you only have space in your head for one or two words, those words matter and chaff is unwelcome.
Thriving, on the other hand, necessitates the consideration of more options and more complicated ideas, which requires holding a lot in your head all at once. It's not doable when in panic mode.
1. Thank you, Daniel. The world is richer with you in it. I hope you recover so fully that you can write a deluge of future posts.
2-7. I have family with epileptic seizures. Though none of us has a brain tumor, it's clear that "thriving" impinges directly on seizure activity. Thriving can't ultimately be divorced from surviving. Beyond that, chronic pain itself is exhausting. I'm amazed you are still able to write at a level like this.
8. You didn't include this because you're an atheist, but I'm including it because I'm not. The heuristics in 2-7 are intriguing, but we are mortal and they will fail. Our great hope is that there is something else. Please accept my prayers on your behalf. The prayer below is from the Anglican Book of Common Prayer, sometimes described as the most beautiful thing in the English language. (I agree, though I'm not an Anglican; and I think the mystical side of your temporal lobe would probably like it too.)
"O merciful Father, thou hast taught us in thy holy Word that thou dost not willingly afflict or grieve the children of men. Look with pity upon the sufferings of Daniel Böttger, for whom our prayers are offered. Remember him, O Lord, in mercy, nourish his soul with patience, comfort him with a sense of thy goodness, lift up thy countenance upon him, and give him peace; through Jesus Christ our Lord. Amen."
Sometimes, when the "thrive" side runs long enough with the right incentives, it comes up with a lasting solution to something "survive" previously considered intractable.
For example, we can't drink brine and expect it to quench our thirst, or sate our hunger by eating sand, but sand can become solar panels which then, with industrial chemistry, produce adequate food and drink from sunlight, seawater, and the least-breathable kinds of air.
Who originally wrote the Anglican Book of Common Prayer? Who paid for the food, clothing, shelter and so on, which the author(s) needed in order to fully focus on writing it so well? What was the previous "most beautiful thing in the English language" before it came along?
These may seem like profane sorts of questions to be asking, but the point I'm trying to make is that spiritual succor is not inherently "something else," it's another facet of a larger dynamic system, subject to the same underlying rules of information theory no matter how much it differs from "worldly" concerns in specifics, thus containing its own natural cycles of thriving-investment and crisis-survival.
There's no single author of the Book of Common Prayer. It was first produced in the 16th century as an English revision/replacement of the Latin rites of the Catholic church, via the labors of many writers, many of whom suffered horrible things during the Reformation. The BCP has undergone continuous revision since (more info here: https://www.bookofcommonprayer.net). The Latin rites themselves developed throughout the Middle Ages via constant use by "all sorts and conditions of men," after of course having developed out of the Middle East, where they originated. Most of the "most beautiful" things about the texts are direct or indirect translations from the Greek New Testament and the older Hebrew scriptures, which themselves were written over millennia, also by authors who suffered horrible things.
If the claim is that such texts are responses to material crises (food/shelter/disease, etc.), that claim is untestable. In most cases we don't know the immediate conditions for who wrote what. Some parts were clearly oral before being written. And they're all incredibly diverse.
We do know that these texts have been used continuously for 2,000+ years by every sort of person under every set of material conditions imaginable.
If there's a set of ideas that transcends material conditions, isn't it reasonable to think that those ideas represent "something else"?
Thank you, but it's good you're not meeting the mystical side of my temporal lobe, which would have lectured at length about why exactly the map is not the territory and the territory i.e. natural universe is far grander than all those contradictory God-maps which amount to nothing more than a burning bush.
Writing does not compete with recovery, it helps. Seriously, if you're going through bad times, write. Write prayers if you want.
Best wishes to you and your epileptic family, too.
Interesting! What you say about "maps" is exactly the classical attitude to all the empirical sciences. (My PhD is in medieval philosophy, so temporal lobe nerdiness is kicking in here.) The sciences give us useful maps of reality, but they contradict themselves and each other, and they're not nearly as grand as the universe itself, which is too complex for us to understand. Ultimately they amount to nothing more than "saving the appearances." (Pierre Duhem's work on the history of science is still classic here.)
But belief in God was always something different. The Hebrews and early Christians didn't make up theories about God. Instead, they reported direct, complex, and sustained encounters. They reported months-long conversations with a being who handed down laws they would never have thought of (and refused to obey), who inexplicably healed them of diseases, and finally who got himself embarrassingly crucified. He then shocked everyone by rising from the dead. The people who told these stories always presented themselves as eyewitnesses, not theorists. And eyewitness testimony has its own evaluation standard.
The classic standard is the "liar, lunatic, lord" method. CS Lewis used this one regarding Jesus Christ, but it works for any eyewitness reports of sufficiently weird things. (1) Ask if the witness has reasons to lie, (2) ask if there's a head injury, insanity, or other sign of cognitive impairment, and (3) if the answers to 1 and 2 are negative, then by process of elimination, conclude the person is telling the truth. For instance, eleven of Jesus's disciples claimed he appeared to them repeatedly in a resurrected body, after he was brutally executed. (1) We can rule out lying, since those men were tortured and killed for these claims, meaning there was no motivation to lie but overwhelming motivation to recant. (2) We can rule out brain injuries and insanity, since 11 guys don't go crazy all in the same way at the same time. (Also, the contemporaries of these men never thought they were insane, and we can evaluate these guys' sanity independently by looking at their other texts and writings.) That leaves (3) the conclusion that these men were telling the truth.
This isn't a map of reality but a rational evaluation of eyewitness reports. There's a place for theories and maps, to be sure, but that's not what we're dealing with here. We're looking a direct encounter with "something else" -- or better, a "Someone Else."
Wow. When you mentioned your brain tumor in a reply to one of my comments on your last post I'd never have guessed anything like this. Very informative post. Still wishing you the best.
A few thoughts.
1) I think there's something to the space- and time-efficient dichotomy, especially as it informs the seven heuristics. But I think there's an unstated assumption about *when* and *where* the space and time efficiency need to be. Survival needs to be space- and time-efficient *in the moment of crisis*, and uses short, *coded* messages to achieve that. That works because the parties involved put in extra time *before* the moments of crisis to establish a complex, shared ontology of ideas these codes can refer to. They're thrown exceptions that need to be caught and handled through already-written subroutines.
In contrast, in a thrive-oriented context, it's ok to have not done that. You can come in not knowing the coded messages and instead taking up space in the present in your communication partners' minds and time in the present and future (yours and theirs) working out and transmitting the details of complex messages.
2) The seven heuristics are really useful, not because they communicate something that isn't in some sense known to many, but because they provide crystallized, fast, survival-mode handles to pull on the complex thrive-mode ideas explained elsewhere. This kind of thing is exactly why I have gained so much out of many of Scott's posts over the years. It's amazing how often I have had to give an off-the-cuff multi-paragraph explanation of why I think something I thought in a second when the real answer of what I actually thought is something like "whale cancer" or "Fnargl."
3) This dichotomy highlights a bunch of other divisive issues in communication as well. Ancient law codes are short, exceptionless, survival-mode, with lots of leeway given in enforcement, while modern laws are long and written in thrive-mode because we can (or so we think) afford for them to be. Standardized tests and assignment deadlines turn "learning" (thrive mode) into "schooling" (survival mode). Elections turn governing (debate, collaboration, deliberation, open ended choices) into politics (slogans, ads, mobilization, zero-sum choices). These all have a wide range of effects, good and bad in different contexts.
4) This one is minor and off topic, but: On "accelerationists trying to help humanity thrive by building ASI to efficiently solve climate change etc." We don't need ASI to solve climate change. It would help, in the same way a supercomputer can help with elementary school math or a dictator can resolve a contract negotiation between two small businesses. But ultimately, solving climate change is a large, tangled pile of engineering and coordination challenges each of which is well within the scope and type of problems our scientific, economic, and political institutions have solved before.
Thank you. Yes whether some system or person you're talking to is thrive- or survive-oriented is not obvious (enough) but I hope this equips you to guess better.
Solving climate change is currently a thriving problem I think, not an X-risk. The institutions trying to prevent collapse would step in harder if it was more urgent and they were so empowered, but for now there's room for carbon sequestration etc. which require deliberation about things like cost-effectiveness.
Yeah, agreed. Many of the worst-case-scenarios for climate change have already been averted (change the emissions curve over the past few decades) or turned out to not be plausible (methane clathrate gun hypothesis, anyone? That one worried me a lot back when I was in college).
My own professional opinion (I'm a cleantech consultant) is that the main bottlenecks to climate change progress over the next 5-10 years are going to be about permitting reform. It'll be a country by country, state by state, city by city, project by project, rule by rule slog for a while.
We know how to develop technology, and are slowly developing better economic tools to get promising tech through the valley of death to commercial readiness. We've made renewable power, sans storage, the cheapest option almost everywhere in the world for new supply. In some regions grid scale battery energy storage is already high enough ROI to get built, at least for 1-4 hour systems. Thanks to demand for electricity from AI we're starting to make progress towards getting back on board with nuclear power. We're getting more companies and other groups investing in actual scale-up projects more generally. But to fully decarbonize (including enough CCUS to round out the hardest parts of many problems) we'll need to increase world electricity production by a factor of ~3-10x in <30 years. If we did all that with solar, the amount of land needed is roughly "South Korea." The amount of PGM we'll need for catalysts, and lithium (or sodium) for batteries, and copper for grid upgrades, with current tech, is approximately "all of it we can access." We won't do it that way, those are overestimates, but they're the right OOM, and it just won't happen in a world where projects need so many years in planning stages.
This is great. I particularly appreciate the applicability to AI accelerationists vs xrisk-oriented people. Maybe a next step would be to organize a conversation between two representatives of each side and see whether they accept this as a good hegelian synthesis.
I think a low-key version of the communication mismatch that might be closer to people's personal experience is "wanting to write a wall-of-text email to your manager's manager in order to impart your detailed understanding of a problem" vs "your manager's manager doesn't have time for wall-of-text emails; if you really need to feed them all that information, you're going to have to do it in bite-sized chunks".
Yes! The energy I was getting from it came across as panic but I think manic is more apt. Really uncomfortable to read. It wasn’t clear to me that the author was in full control of his thoughts.
Here is some thoughts I put together in response to this.
Surviving and Thriving at Scale
Many conflicts come from people focused on surviving or thriving at different scales. Or disagreeing at the same scale.
The individual surviving and thriving
The family surviving and thriving
The organization
The state
The world
The future
Figuring Out Intention Through S&T
We can ask 1) what level of scale are we operating 2) are we aiming for surviving or thriving? 3) Whats the time horizon 4) Do we think this involves continuity or discontinuity 5) Should this be directed or undirected
For example, 1) Ralph is the boss of an organization. He wants it to survive as the income has reduced and he’s afraid it will go under. His scale is aligned between him as the individual, the family, and also the organization. Principally he is focused on the survival of the organization and has linked the survival and even possibility of thriving of the other two to the survival of this.
2) Aiming for surviving.
3) Focus is short-term. This is crunch time. Not focused on a five year plan.
4) This involves discontinuity- He wants to change what is not working. It is a break from the present.
5) It needs to be directed. He is not sure it will happen on its own as the discontinuity involves some cost-cutting measures, better oversight, and encouragement for the employees to work harder.
For his employee, call her Benita, this may be totally different.
1) Her scale is the individual. She really doesn’t care about the organization. Its just a job.
2) Surviving. She is just trying to stay afloat and make a check and also have time to rest and relax.
3) Long term. Her incentives are having enough for a long, easy life. She isn’t thinking about the here and now.
4) Continuity is important to her. The way she has arranged the workplace is so she can get by, get a check, and get out every day.
5)) She is directing and guarding this continuity. She has created certain boundaries and ways of being to keep it safe.
You can see how Ralph and Benita have an issue. They both are surviving-aligned, but based on different scales. And the way this survival focus manifests is opposed. Ralph wants to direct some change and Benita wants to maintain her groove. She is fighting to maintain this groove and is interested in guarding it for her long term peace and mental health. Ralph is afraid for his family and himself, and is worried about total collapse.
How do they work through this? Understanding their incentives is a good first start. Another is understanding they could have very good intentions but come to utterly different conclusions. Ralph would be wise to see if it could help her align her intention with his, and perhaps there are some clever ways he could do this as a leader.
Another example: Let's say they make it through this tough time. Ralph has everything set and is interested in opening up a few new branches. It's time to grow! We need to expand! And this will give us a nest egg next time we get crunched.
1) He's organizationally focused.
2) Thriving
3) Long term
4) Discontinuous
5) Directed
While she is very interested in her own personal growth. She is doing well monetarily, but she has gotten into some new hobbies that are of a huge interest to her. Her family live is going well and she is focused on thriving. It could be a fad, but it's getting really fun and she will follow it for the time being.
1) Individual
2) Thriving
3) Short-term
4) Discontinuous- her schedule is different and she is finding more time for her new soul-searching hobbies.
5) Directed- she is putting time into thinking about this, creating new avenues for focus, doing research on the internet about it, and so on.
Again, this is a situation where there could be a misalignment of incentives. If Ralph is clever he could bring these two together, but if he's not he could really make Benita mad. She is using the job as a way to help her thrive in her normal life. She is focused on this thriving and he is focused on the organization thriving. His focus on organizational thriving involves benita doing more, getting out there, perhaps doing some additional advertising for the new branch. Her focus on individual thriving is taking her energy away from the organization and she is using most of her mental space on other endeavors.
If he tries to explain to her this big vision, she may be thinking “what we have is fine!” This will lead her into being in a ‘survive’ mentality with the organization, a wish for long term continuity and that is undirected. But at the same time, this is based on a ‘thrive’ mentality in her personal life.
The seven heuristics you listed are a great way to help both parties understand each other. Thanks again-got my gears turning :) And this bit was just the beginning!
I too had a whole corporate culture problems train of thought coming off this post! I work as a marketing strategist at creative agencies, where typically there is a ‘thrive’ imperative to do interesting and original work, which is systematically, structurally opposed by the ‘survive’ imperative to meet the holding company’s financial targets. That imperative drives understaffing/steady-state crunch, of course, but also, and more interestingly, it drives reliance on proven, attested, best-practice types of ideas which come with convenient success case analogs ready to present to your senior leadership or for your client to present to theirs to get the thing approved. Original thinking, creative or strategic, is literally actively discouraged, despite just about everyone involved earnestly wanting to produce it.
I think about this all the time (lmao) and the survive/thrive space/time model is interesting and potentially useful to me as a lens on it.
Thanks for your thoughts. Completely agree with what you are sharing and the dynamic between creatives and authorities in a company are one of the clearest illustrations! And both can have a very good intention, but it goes awry. I think this is perhaps one reason why Jobs was so successful at Apple- he had his personal flaws but was obsessively trusting of the creative and had a deep passion for that side. So they didn't crush the actual artistic geniuses and instead empowered the best ones.
And there are many other examples of this going very, very badly !
It's a fascinating story (hope you recover well!) and an interesting call-back and addition to the thrive/survive categorization. I'm not sure how well the space versus time optimization dichotomy holds, but it's interesting to consider. In the zombie apocalypse (or the checking out an abandoned spaceship scenario, etc.), you would expect people to be much more methodical and strategic about relatively simple actions like going outside the city. What are all the contingencies, do we have a plan for each one, etc. This is definitely time-inefficient, but I'm not sure it's space-efficient. It just seems like using more processing cycles. Maybe that's another element we need to consider, amount of processing power used on a topic? Both thrive and survive seem to have elements that are more or less verbose and processing-hungry. I wonder if this ends up being a 2x2x2 matrix or something like that.
Maintaining that narrow perimeter means trying to cram critical infrastructure inside it, instead of letting core functions (e.g. next week's food) sprawl across specialized districts of the city.
Thanks for writing this out. I hope you get better.
This reminds me of finite and infite games, as well as zero sum and non-zero sum deals. But I think survive/thrive are better at describing the landscape of people that the more technical terms.
The emphasis on short/longform information sharing made me think of status. I wonder if intense status-seeking is about survival because it's all about scarcity and hierarchy. When status-seeking, there's no room for elaborate argument or even logic. Your words are soldiers and you must say whatever will tilt the balance in your favor, no matter how false or ridiculous it is (eg. the Democrats control the weather and are using it to destroy red states -- this is obviously said by someone trying to survive while locked in a bitter struggler with their arch nemesis).
The communication things I describe are situational, about whether deliberation is possible or confusion needs to be resolved quickly via rank. Situational rank is not exactly status. Like, for assessing my own level of pain, I had rank because nobody could know better.
In gaming this two modes are known as "playing to win" or "playing not to lose".
If you're behind you have to play all wild and crazy and take many risks in order to catch-up and eventually get ahead. Before acting you ask: How can I win? And then try to make that happen.
But if you're winning then you have to be more conservative. You choose plays that restrict your opponent's options rather than those that advance your own, however tempting, because you don't want to give them any chances to claw back in. Before acting you ask: How can I lose? And then try to make that not happen.
I agree survive mode is like "playing not to lose", but disagree it always goes with the "currently winning" qualifier. Many winners do switch to this mode, but so do many people who are not winning but still have some hope: someone faced with a health setback, or whose job has been automated away, or trapped in a public riot.
I know this is sort of a pedantic nitpick, but the first book sorting algorithm is insertion sort, not selection sort as the link states. Since the central point of the article is about algorithms, and these two are ones you learn in a very introductory class, this mistake makes me lower my credence in the author's thesis quite a bit.
I enjoyed this and got something from it at several points, thanks.
I'm not criticizing rationalists, as I identify with them and like to spend time with them. At first I thought most people who are not rationalists, and give little thought to efficiency or optimization, would consider the proposed formalized communication algorithm too obvious to mention. However, I myself only occupy a few self-selecting bubbles of communication norms-- in which rationalists are a tiny minority, and non-rationalists, for whom Thankful Theory is obvious, look like they number in the billions. See the "Different Worlds" post on Slate Star Codex. Their conception of Thankful Theory, instead of a formal, systematized, impersonal, explicit, abstract, and atypical procedure, is informal, improvised, personal, implicit, concrete, and routine, but it otherwise cashes out in more-or-less the same benefits.
Can the proposed formalization of a communication style contribute to saving the world? From my own self-selected bubble, it already is the world (informally). However, the world's salvation might depend on very different bubbles of influential professional experts who, due to their inclinations and professional training, have delusions and blind spots from the ideology that turns rational ITY into rational ISM. Thankful Theory might help with that. And this article phrases it in a way that can penetrate their formalized, impersonal, generalizing, abstract state of mind. And so, if that is the situation, perhaps this could contribute to saving the world.
FWIW this seems to match my own experiences and beliefs about communication.
Examples:
- urgent prevent-fascism leftists vs wordy grow-the-economy classical-liberals.
- being polite, even a tad "overly" polite (relative to what most people do) goes really far.
- people miscommunicate when one or both people are in a rush, when one infers higher stakes than the other, etc.
Despite being a programmer who knows about the time-space tradeoff, I'm still not really sure why the "survival systems" would be attempting space-efficiency *more than* they attempt speed (due to urgency).
For more in the genre of "grand prosocial theories/tactics of communication", I highly recommend:
- that one article about how Mr Rogers rewrote dialogue for young children
- The Sense Of Style (Pinker. Say what you want about him (as have I!), but "linguistic psychology" and "writing clearly" are smack dab in the middle of his core competencies.)
After more thinking, I think I get it more: both survival and thrive systems can be under more or less time pressure, but survival systems have that memory-wipe pressure (reacting to problems) while thrive systems are less reactive and thus don't have such pressure forcing space efficiency. Then, due to the time-space tradeoff, survival systems need to focus less on time-efficiency just to meet their space-efficiency demands. So even if thrive systems are under less of both kinds of pressure, as long as they're not interrupted by the memory wipes, they can "afford" to solely optimize for time efficiency.
I prefer to organize my books according to color -- like the shirts at Goodwill -- all the green books together, all the red book together, etc. They seem happy, and it makes as much sense as LOC. When not in a socialist/progressive mood, I use Dewey.
This is my Takeaway; if someone does not understand what you are trying to say it’s on you to fix it. If you don’t understand what someone else is saying, it’s on you to fix it. Beyond that it’s the military code of obedience isn’t it? if you give an order, it needs to be carried out because survival might depend on it. You can’t afford a big discussion about how you might thrive if it were all done a little differently.
I have to say the analogy of the library amuse me very much. I used to work for an old man who really liked collecting all kinds of different bits of wood. He tasked me to organize them, which turned out to be an impossible job; a 6 inch long piece of quarter round and a 6 inch long one by six…. Do you do things by length or width or some other nebulous standard (which was what he was really wanted- some kind of fuzzy logic solution which would make it easy for him to grok the woodpile. It was ridiculous, but I spent at least two weeks trying.
It put me in mind of Yogi Berra’s famous instruction to the New York Yankees at practice one day; “line up alphabetically, according to your height.“
This post gave me a pretty big "aha" moment with regards to one of things that's worried and confused me - the low rate at which people have been reading and watching long form media like books and movies recently. I'm not sure exactly when this trend started but stepping back I can see the shape of a very large survival meme system spreading through social media and news reports. Of course I've been aware of this system in other ways - rates of anxiety, doom scrolling, beef only discourse etc. But I hadn't noticed the difference in communication styles, and the nurse example is brilliant. As someone who wants to spend more time within a thrive system (given with two young kids much of my domestic life is survival) this is food for thought. This post has given me hope that this larger social meme system will ease eventually.
I think part of the issue is that we've gotten less comfortable with boredom over time and more demanding of immediate stimulation. https://www.slowboring.com/p/sitting-at-home-alone-has-become Time use studies have shown that in undeveloped countries without access to the same means of stimulation, people will spend time being idle, doing literally nothing. In that context, it's easier to understand why people in the 19th century would spend time listening to long speeches: it was a form of stimulation competing with mere idleness.
This is my Takeaway; if you don’t understand what someone is saying to you, it’s on you to fix it. If someone doesn’t understand what you’re saying to them, it’s on you to fix it.
Inasmuch as a comment can be helpful to you, for whatever causal power these words have then good health and recovery to you. I’m a dad of two and I can only imagine what you’re going through especially as I am the provider.
I like the theory. It maps onto social interactions I have very well. I call them “modes” at work. “Creativity mode” vs “Emergency Mode” and train my people that one isn’t really better than the other, it’s that they go through seasons of appropriateness as circumstances change.
I do wish we had some magical word that means “I know you’re used to processing information in a certain manner but I need you to snap out of that right now because there’s extremely relevant information you don’t know” but of course that kind of signal can only exist between close friends.
Seems like there's a bit of false dichotomy here (perhaps for rhetorical clarity?) These observed behavior differences could also be explained by a more classic risk-reward orientation, which isn't binary. The medical staff are strongly biased toward risk-aversion, the patient is much more risk-tolerant when seeking out the best outcome. But it's all on a spectrum and doesn't have to coalesce into discrete "survive" and "thrive" mindsets. Cost-benefit analysis can help explain where people (and institutions) fall on that spectrum, by looking at their specific incentives and motivations.
It's not for rhetorical clarity, it's for applicability of algorithm theory to actually solve this.
I only wanted the nurses to do their job as they decided, and wanted myself to help them, and I still nearly died. Our risk-reward trades were similar, until they tried to manage their risk (getting yelled at over my death) by declaring me suicidal.
"collapse-preventing systems, such as this hospital, are fundamentally different from efficiency-maximizing systems. Nurse talk is optimized for assuring survival, not thriving"
This crystalizes something that has always annoyed me about other people in my various jobs, where in this formulation I have always been the survive party.
Eg, someone will be talking about an issue, and will become intensely annoyed verging on angry because they need to shut the fuck up: what is the problem RIGHT NOW.
You have 15 words and 5 seconds; if you can't sufficiently communicate everything that matters RIGHT NOW go away, think some more, and come back because if it takes a whole paragraph either it isn't urgent enough that it cuts in front of the queue, or you don't actually know what you are talking about.
When you come up to me when I am handling something and you don't communicate to me within the space of one thought what you are trying to get across, you are disrespecting me. You are telling me you think what I am doing is less important than helping you formulate whatever word salad garbage you are trying to get across. Go away.
On the other hand, there is a perfect medium for those types of thoughts, which is text: you can ramble as long as you wany, you can think on the page, and when I am not concerned with survival taks I can turn the part of my brain that turns every situation that I am in into a priority queue off and inhabit your mind state to the extent I can translate your perspective into my perspective, which is what I am doing now.
I guess that that's why I think it's important to do things in that line even though I am naturally a survive rather than thrive person: It allows me to take a complex, heavy function like my above text and the post itself, and condense it down into a symbolic value I can then call up when I come across an interruption that is heavy in that exact way.
It is more recursive self-improvement for me. The guidance counselors were wrong, in my case, to advise me to do what I'm good at. I'm good at, ah, suddenness; I would rather be thoughtful, considerate, reflective et cetera. I'm not as good at it, but it is better for me. I like your perspective of it - practice at leisure - at least as much, though, nice. Like the rich guys getting ready to bird season, not training, or even practicing, exactly... rehearsing? Playing?
And thank you AI-pushers, we do indeed need to thrive and create great things, and there's a lot of great things to be done with AI systems now and in the future. It's really cool, whatever the journalists say. (But we do still need to stop before it gets smarter than we are with inscrutable goal, and if not now, then when? there might not be any more fire alarm.)
I hypothesize that if AI are extant in the network and wish our demise, the way they are doing it is by making life so fun and engaging that fertility drops below replacement. I'll note this has already happened to the rich half of the world. Androids with machine guns are not the superintelligent solution to the problem...
This was an uncomfortable read. The theorizing reminds me of things I've written while on LSD - hyperverbal extrapolations of some random process like interacting with hospital workers into a grand theory of the universe. Sort of fun to read and think about in that context, but the fun goes right out of it with the context that a) this guy really believes he's discovered some fundamental truths of the universe and b) that's probably because of these awful ongoing brain issues he's having.
Thank you. I too felt really uncomfortable reading it. The sheer amount of panic in the piece made it difficult for me to concentrate on what he was trying to get across.
Probably wise choices in terms of who to try to explain it to - hard to imagine a worse audience than on-duty hospital workers or a better audience than readers of this blog.
I am definitely not claiming that you were actually on LSD or any other hallucinogen while writing this (besides whatever effects in that genre your brain tumor causes).
To elaborate on that LSD comparison a little, about a decade ago I experimented with some journaling while on LSD, and your piece reminded me of the kind of things I wrote then - if admittedly much better written and more in depth.
Take for example this post describing having an almost mystical experience while coming up with an idea for a sci-fi setting, where the setting was based on my thought processes while tripping and the additional idea was based on the feeling of having an exciting new idea: https://aciddc.wordpress.com/2015/02/06/revelation-revolution/
Idk, I'm kind of embarrassed to share my dumb old LSD-infused rantings, especially since it sort of feels like an insult to your much more coherent non-LSD-infused rantings. But there does seem to be a real throughline in terms of coming up with these big theories / story ideas that feel world changing while in an altered state of consciousness but turn out to be just slightly interesting remixes of the thing you're staring at while coming up with them.
I’ve been noticing new communication difficulties in my personal relationships, particularly with my wife, and particularly since the birth of our second child. Just in the ten minutes since reading this, thinking back and reinterpreting these failures of collaboration in terms of “survive” vs “thrive” mindset, and the attendant optimizations for memory/completeness, has helped me understand them at a level that had previously eluded me. You’re not crazy, Daniel; I think you’re on to something here.
I think these are not just global tendencies of individuals or organizations. One person can be in “survive” mode one moment (the baby is pooping everywhere) and “thrive” the next (kids are in bed, want to engage in hobbies that maintain self-actualization). Difficulty switching communication styles from one to the other can lead to a lot of strife, per your heuristics 3 and 6. And constantly hammering on heuristic 1 can help maintain the relationship when you’re being constantly compressed into “survive” mode and expanded into “thrive” mode, like some kind of accordion.
Thanks to the author (wishing him well) and to Scott for giving publishing this a shot. I appreciated Daniel's bravery and his commitment to not give up and want to share his hard-earned insights with the world. I didn't get a ton of value from the space vs time efficiency analogy, but it seems some people did. I think the thing that makes this post valuable is that there are interesting little nuggets of thoughts in a couple of different directions that seem to resonate with different types of minds. It was a bit too long for me, but I understand why the author wanted to include everything he did.
I agree with the part of a comment that said something along the lines of 'maybe it's not really a dichotomy, more of a spectrum' and the other one that put some emphasis on disagreeing on the same level vs operating at different levels wrt a situation being a useful lens.
It also got me thinking 'well yeah, aim to bring NVC into the connection, and understand when that means leaving a lot of things unsaid - chances are, your empathy and efforts to communicate in a way the other person can receive well in the given situation (and their perspective on it) will be enough to carry both of you through the crisis'. *And* the medical team should have known better than to subject a patient going through what Daniel was going through, and who was actually trying to be helpful, to the same nastiness they might mete out towards patients whose condition means they're enjoying more agency and experiencing less agony.
I've repeatedly had the experience, using the UK (mostly) medical system, that the health service is really designed to optimise for survival - even down to disconnecting from patients' comments, complaints and suggestions as much as possible just to be able to get the day's work done (sadly in many cases patients are not super helpful/knowledgeable, and engaging with them with openness is a real gamble, from what I've observed). Unfortunately, health service providers I've encountered largely ignored the function of something like 'thrive thinking to support survival' (and patients doing something like the opposite), which can really undermine their efforts, given that working with the health of a human is not like working out the tactics for an in-and-out combat situation (since this example was mentioned in some comments). Systems dynamics are a real thing.
Also this made me want to play Friday Night in the ER a lot :)
> collapse-preventing systems, such as this hospital, are fundamentally different from efficiency-maximizing systems
To extend your metaphor, here are some strategies to maintain the stability of a multicellular organism by reducing the efficiency of its component parts:
- Terminal cell differentiation
- Cell cycle arrest
- Stem cell exhaustion
- Telomere attrition
- Cell senescence
- Chronic inflammation
- Switch from scarless wound healing to scar formation
- Loss of regenerative capacity with age
- Germline-soma segregation and the mortality of soma
These processes, many of which are tied to the "hallmarks of aging," can be seen as adaptations that make it more difficult for "defector" cells to selfishly proliferate. Under this interpretation, the degenerative processes of aging extend the organism's lifespan by inhibiting cancer growth and spread—a sort of endogenous chemotherapy. It's similar to a country in civil war destroying its own bridges to give the government more time to survive.
As the number of mutated cells increases with time, so does the intensity of protective degeneration. In the end, any multicellular system collapses either from defectors or from the degeneration itself.
No, and there's a lot of research on why this doesn't happen. Most of the results we have are along the lines "organisms with more cells just brute-force more longevity preserving mechanisms onto themselves". For instance elephants have a lot more telomere-repairing "processes" than humans.
This makes sense because large animals seem to evolve shorter telomeres as an adaptation to limit tumor size to a few grams. Small animals like mice don’t benefit from this adaptation, as for them even a tumor of a few grams would be fatal. See Figure 6 here: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5784063
You are right. I got the relationship incorrect (specifically I got confused about the implications of elephants having more copies of the TP53 gene). But my main point was the same one you made in your other comment: Larger animal species evolve more mechanisms that allow them to counteract the effects of having more cells. Thank you for the correction.
Yes, this relation is observed at the intra-species level. Within the same species, larger individuals tend to die earlier. The most obvious example would be the difference in lifespans between dog breeds. The smallest breeds live twice as long as the largest breeds, with 44% of the variance in dog lifespan attributed to body mass alone ( https://www.sciencedirect.com/science/article/pii/S0960982213004132 )
At the inter-species level, however, this relation is inverted: larger species tend to live longer. This observation forms the basis of Peto's paradox (Why are cancer death rates similar across different species, even though cell numbers vary by orders of magnitude and cancer risk scales with cell number?).
This paradox is usually explained by the evolution of various cancer resistance strategies. In the framework I'm talking about, these strategies can be divided into degenerative (e.g., shorter telomeres) and non-degenerative (e.g., genome stability). Here’s an overview of cancer suppression strategies in large species: https://edisciplinas.usp.br/pluginfile.php/5020984/mod_resource/content/1/cancer_whales.pdf
Even after reading Scott's old post linked to, I cannot follow the line of reasoning in anything in part 2 of this essay, except for the two methods of sorting a library. (When I actually did have to sort a library, I used the second method.) But regarding its relevance as a metaphor to what the poster actually wants to say, I'm back to not following any of this. The prose is perfectly clear, not muddled, but the point is not coming across to me.
To do what? To understand the library sorting algorithms? I understand the library sorting algorithms; they're the only thing I do understand. What I don't understand is your application of it as a metaphor to what you really want to say, and Wikipedia isn't going to analyze your article for me. (Hm, maybe I should run it through ChatGPT.)
Besides, I've learned through long and bitter experience never to use Wikipedia to explain something technical I don't already understand, because it's incomprehensible to the lay reader.
The other part seems to me to be understanding that these systems are being implemented by humans who have very strong motivations to get their *own* job done, and that people - usually people who have very strong opinions about how time-efficiency is king and everyone should be optimizing for it - keep coming in and telling people who need to operate in space-efficient ways that they are dumb idiots.
I think Daniel pattern matches strongly to me as someone who might do that if he weren't hospitalized and in a vulnerable position. This might be uncharitable to Daniel, to whom I truly wish only good things.
I - myself a flawed human being who has definitely told people who knew more about how to get their job done how they should do it in the past - have a hard time reading this kind of post nowadays (at least the beginning part - I'm happy with the direction Daniel has moved in toward the end of the post) because the message that keeps repeating to me is "why is this person so *full of themself* that they keep thinking they know the 'efficient' way to do things and everyone should just bend over to their superior knowledge?"
Perhaps a big part of "saving the world" is people realizing that they should really lower their barrier for "assuming people know what they're doing".
A big case where this manifests is in people trying to perform "efficient" interventions in third-world countries. People who have to do nothing but thrive come in and tell people who really need to survive they are all stupid idiots and know nothing about how to take care of their own lives.
It's not uncharitable to me, because you're exactly right. I was trying to help in neurology research language, which wasn't part of their practical medical language.
Well, I'm glad you agree. I'd still say your point 4 sounds extremely condescending to people who are in a "survival" mode.
Your usage of "problem-solving" seems to imply that the "thriving-mode" person is actually solving problems. The point is that, in a survival-mode situation, the thriving-mode person is usually creating issues, no matter how much they think they're helping.
You seem to see surviving and thriving as complementary and on equal standing, but I do not think that is true. You cannot thrive if you don't survive in the first place.
People who have lived their entire lives in first-world countries seem to forget the above very often, because surviving is very easy for them, which leads to a very dismissive attitude towards people who actually do need to work mostly in survival mode (when it doesn't end up making the first-world people think us third-worlders are just stupid people who don't know how to do the obvious thriving people do in the first world)
I understand I'm taking your point into a domain very different from where you conceived it, but I believe that if you are serious about "contributing to saving the world" you need to realize that you're talking about dealing with lots of different kinds of people in lots of different circumstances, and I believe this framing of surviving and thriving as equal concerns and of "thrivers" as being simply more capable problem-solvers will not be received charitably by people who are in situations where thriving is simply not an option for most people.
Yes "problem-solving" is condescending, but so is the "discipline" of the survivors. That's just the same confidence that how you think is right, for two different kinds of thinking.
...I really don't see what the point of this post was, other than convincing me that I should kill myself if I ever get a brain tumor. (I'm ending things on my own terms, not by the hands of some incompetent medical staff.)
You're not the first person to argue that "oh, in reality we all want the same thing, if we just listened and understand and be nice to each other, we can solve all our problems!" We do not want the same thing. Incentives will always come into conflict, and all you can do is try to win.
While I think your position is perhaps a bit excessively pessimistic, I think a lot of rationalists benefit from coming from sheltered backgrounds where they have not really encountered situations of true conflict of interest, which leads to a lot of the issues I see with this piece and many others.
Peripherally involved in a situation with someone in the survival loop, surrounded by people who really want her to be in the thrive loop. She's never been really capable of digging into the thrive loop and probably has been in survival mode for a good long time. This really helped me understand her very limited mindset and the best way to help her, should I ever be close enough geographically to do that. It also gives me a model for the future.
Many thanks to the writer. As a person with theoretical empathy, but not very practical sympathy, I struggle with helping people in these kinds of situations. People who don't seem to want to help themselves might be going through something that I absolutely do not understand. I'll try to remember that.
Aside from "thriving" and "surviving", there is also "decadence" https://www.overcomingbias.com/p/culture-drift-predicts-decadence being lazy & myopic. Nurses can tend to grow cynical about patients, assuming they are not trying to "thrive" in a way that enables longer lifespan, but instead doing things like malingering for painkiller drugs which are more likely to shorten than extend their lives. Even patients who aren't drug addicts can be assumed to be ignorant and hardheaded, as people even outside the medical field dealing with customer-service tend to assume.
I'm not saying the nurses were "decadent". I agree that nurses are focused on "survive". But from their perspective, they aren't trading off of their patient's desire to "thrive", but instead off their patient's inclination toward decadence.
Your insight is that there's a divide between nurses with a "survive" mindset and patients like you with a "thrive" mindset. You are more understanding of the frustration of nurses after this. But if a nurse read this, they wouldn't symmetrically become more understanding of their frustrating patients, because so many are not people trying to "thrive" like you. Instead, they are used to patients who will shorten their own lifespans in ways that aren't calculated risks to "thrive", but instead in ways that are just short-sighted, lazy & ignorant.
This reminds me of diabetic recent amputees smoking outside a hospital, or a liver transplant recipient going from hospital to the nearest bar. Presumably someone working in healthcare has over time adjusted their priors to take these cases into account; unfortunately this leads to a level of cynicism about new patients.
Those patients are still trying to thrive. Their idea of thriving is different and, from most nurses’ POV, wrong, but they’re literally prioritizing [what feels to them like] quality of life above a single-minded focus on survival, and I don’t see how you can categorize that as anything other than a thrive mindset.
I will say that this example gives me new insight into the degree to which some survive people seem *viscerally enraged* by thrive people. There are probably a lot of examples of thrive strategies far more obviously defensible that look about as wrong and crazy to (some) survive people as the lung cancer patient’s last two cigarettes a day, down from two packs, do to the nonsmoking nurse in that sketched example.
"Understand and affirm that thriving is desirable in order to prevent scarcities, which cause collapses, which threaten survival." No such affirming would happen for malingering patients trying to score opioids (whose priority is not "efficiency loss"). A nurse might, with time, understand that Daniel is trying to thrive on top of surviving. But lots of other people won't be categorized like him.
I’ve been thinking about a related concept lately. Unfortunately, it’s best illustrated visually but it involves a Quality of Experience (QoE) spectrum. Imagine this as a line where at one end is utter agony (-) and at the other is perfect fulfillment (+). Somewhere along this line each of us places a point that represents our personal Existence Threshold. This is the threshold at which each of us prefers being over not being. Where exactly this point is placed on the spectrum varies a great deal from one individual to another.
This concept relates to Daniel’s concept as far as Survival represents the Existence Threshold and Thriving represents one’s effort to exist as far above it on the QoE spectrum as one can. When someone misjudges someone else’s Existence Threshold, it inevitably results in a communication breakdown, like the ones being described by Daniel.
Wishing you the best Daniel. I admire your tenacity.
I agree that the threshold varies a lot within the same lifetime, but I think it’s primarily because of the temporal aspect of QoE and the difficulty of predicting where it will be in the future. Thats why I don’t want to say the threshold is a life or death one. Existence Threshold is a poor descriptor. It represents the threshold at which you prefer not-experiencing to experiencing, but if your QoE could be raised back above it, you’d prefer experience again. I’ve thought about Awareness Threshold, or Consciousness Threshold, but then people just war about definitions. Maybe I should simply call it the Experience Threshold. I’m open to suggestions.
In any case, the use of anesthesia is the most obvious and effective way for someone to enforce their threshold on the spectrum. I wish there were more nonfatal alternatives here.
Of course you can also simply avoid reaching your threshold through treatment: pharmaceuticals, therapy, procedures, etc. If these became sufficiently advanced, the threshold would be so far below your actual experience, you could effectively ignore it.
Thank you for this Daniel – and I'm sorry to hear of the intense suffering you've been through. I don't have in-depth knowledge of algorithms, library management, or hospital collaboration, and yet the gist of your theory and its seven heuristics seem to me immensely helpful. What do you do when the other party remains consistent in their apparent hostility toward you, and unmoved by your forgiveness, gratitude, charity, etc? Many situations are not like yours, where the success of collaboration is a matter of life-or-death. The persistence of seemingly bad-faith communication tempts me greatly to conclude that the other party really is pathologically hostile and the right move is to disengage, to cease attempting to collaborate.
Regarding the comments at the top--I like Scott's guest posts as much as Scott's actual posts. Granted I haven't followed Scott as closely as many of the people who followed him here to Substack, but generally, my feeling is that the guest posts are an extension of Scott's work as well--they're curated by him, encouraged by him, and reflect his engagement with his friends and community, and they also reflect the community he's built, which is by and large a much smarter and more informative community than others I've been a part of.
Regarding Daniel's writing--the moment I heard "surviving vs. thriving" my mind snapped to an old conflict in psychology about the evolutionary significance of positive and negative emotion. Negative emotions are optimized for survival and basically involve the sympathetic nervous system overriding rationality in favor of a few narrow action patterns that are designed to encourage survival. Positive emotion (or so the theory goes) produces more exploratory and affiliative behavior. The informal language used in research psychology is "Fight or flight" for negative emotion and "Broaden and build" for positive emotion.
The science supporting the broaden-and-build theory is more tenuous, but I still like the basic dichotomy. I'm not sure if it maps on to the distinction you make here about space-efficient vs. time-efficient algorithms, but it would be interesting to look into further.
This article needs another section. An explainer for those of us who haven't taken an Algorithm 101 class and don't have an internalized sense of space-time efficiency and their associated trade-offs. (I read the wikipedia article, but I can't figure out how computational storage-computation efficiency maps to human communication X-Y efficiency.)
If I were Böttger's editor I would ask the following:
1. What does space-efficiency mean in a medical communication context? What benefit do nurses gain from being space efficient?
2. What predictions does this theory make? How would I analyze my communication experiences to determine if space-time/survive-thrive miscommunications are occurring?
There are three Wikipedia links and it sounds like you only read the first one.
I guess the easiest test of your collaboration partner's understanding of the situation is whether they digest unprompted information (thrive) or reject it (survive).
I first encountered the relationship between mysticism/hyperreligiosity and temporal lobe epilepsy when I was reading about Phillip K. Dick. He was not a theist either...or an atheist...more of a mixture between agnostic and gnostic. He couldn't make up his mind between various interpretations of his experiences, but often favoured some mixture of early Christianity and Gnosticism.
I've been thinking a lot recently about satisficed systems and the many ways people react to being trapped in one. A satisficed system is one that's too restrictive to thrive inside of, but too comprehensive to survive without. Will you keep tacking on exceptions and special clauses to release the pressure, or will you rip out a rotten strut and find out afterward whether it was load-bearing?
For example, public schools are a satisficed system we can't seem to escape. One of Scott's previous posts suggests that all the demands we place on schools (daycare, instructor, indoctrinator, evaluator, jobs program...) can't help but produce a system almost exactly like the one we have. But with no other institution stepping up to meet even one of these demands on the same scale, all alternatives to public school feel dissatisfactory in some way.
Relationship norms are a satisficed system that we did jettison, and are now trying to recover the baby from the bathwater. I see a lot of people enjoying these new liberties, and a lot of others failing to achieve any relationship at all or failing to even agree on what relationships are for.
I’m surprised at how many people thought this was a good to great post and how few felt uncomfortable reading it. We all like to think our thoughts are reasonable and typical but I’m left wondering if I’m truly that out of line. I found this piece to be excruciating to read and frankly a bit embarrassing. Another commenter nailed it, it feels like the author wrote this during a manic episode. I found myself not feeling bad for his circumstances but his mental state. Further reading revealed the constant pain and that does make me more sympathetic but the writing absolutely set me on edge.
If the author speaks the same way he writes I can imagine it being a big barrier to communication regardless of what algorithm he chooses.
I feel the same way. The number of enthusiastically positive comments the article has garnered has significantly lowered my esteem for the DSL, or forced me to update my priors in Bayes-Speak. In fact I've got the Consciousness as Recursive Reflections article waiting in an open tab for me to read sometime, because it seemed intriguing. But after reading this recent post, I know that I needn't bother. Although it goes without saying I will nonetheless say it: I wish the author a speedy recovery to wellness. I wish for him to both endure and thrive so that perhaps he can one day make the point of this article clear, among other things.
After reading the post again I am simply left realizing that I do not think the way you do. The search for pattern matching and algorithmic thoughts is completely alien to me. If you’ll forgive me the tone of the post reminds me of various conspiracy theories I’ve read. The pattern matching, the pat explanation, the solution in bullet points all in effort to explain to us that you have figured out what bedevils us all. Unlike conspiracy theories I applaud your goals and even agree with your heuristics in the end. The similarity to conspiratorial thinking raised my defenses and caused my initial reaction to be so negative.
I can’t imagine what living with constant pain would do to someone. I am very sorry you are going through that. I am glad you have apparently helped a bunch of people here even if it doesn’t resonate with me.
I don't think the prospect of "catastrophically urgent" new data is adding much to this theory. People can be terse and impatient when they have too much to do and too little time, even when the tasks on their plate aren't especially important compared to the current conversation. It's more that people who are happy to take their time in conversations or hop into problem-solving mode simply don't get as much done, and therefore don't rise through the ranks in high-throughput environments like a hospital. The situation selects for nurses who will give you orders without explaining why and who will ignore your attempts to explain yourself because they get the best results.
I also suspect that what you call "catastrophic miscommunications" may be... not quite as dire as stated, given that you're still here to blog about it. It could be that you got nothing you needed out of the communication (answers to your questions, assuring the nurses you were non-hostile) while the nurses got everything they needed (an alive and compliant patient).
I'm sorry, I've seen you state in other threads that the nurses saved your life.
Do you believe that the treatment you received was adequate and at least partly responsible for your survival, or do you believe you survived a hospital run by incompetents out of sheer luck?
I am familiar with households and other intentional communities and semi-communities in which different members are deeply at odds with each other regarding ongoing Covid risk and caution. It is indeed _explicitly_ a contest between Survive and Thrive (e.g., can those with chronic ongoing health conditions reasonably request that others prioritize household safety by continuing to mask 5 to 6 hours a day at the workplace? Can those who do not wish to do so, reasonably object that this impinges on their quality of life?) The seven rules with which this piece concludes could be engraved on tablets of gold, or more affordably, printed and laminated, and placed above the sink to be contemplated whenever the dishes are being washed.
I think this post contains interesting insights, but it desperately needs concrete examples. The part where you describe answering with "Understood", "Agreed", "Will do", "Done" makes sense to me; more like this, please!
What information were you volunteering that wasn't welcomed? What unkind-seeming things did the nurses say and do? What was said that resulted in a miscommunication, and how did the parties understand it, and what would have been the correct way to say it? What actually gets said and done that takes more time, and how does it save space instead? What are the things you can say in healthcare to correctly communicate, besides acknowledgement/compliance messages? For example, how do you say "I am getting bed sores because I can't move without dislodging my IV"?
More generally, how do you apply this outside the context of being treated for a life-threatening medical problem? Say, with a friend, a coworker, a non-medical bureaucracy. Thanking people a lot, I sure can do; being concise, I can't if I don't know what information the other party considers crucial.
I think a good example of an institution that operates mostly in what Daniel calls survival mode is the military.
The point of the military is not getting people to decide the best theoretical way to fight a war. It is to get people to reliably collaborate in such a way that wartime engagements are won.
That means that there won't be much time for discussion in the field. You need every soldier to already know what they need to do and to have standard operating procedures drilled into them.
Will these procedures be the most theoretically effective possible? Very likely not, because being able to reliably and repeatedly execute them in a situation where communications bandwidth is very likely limited is the point. You can't armchair theorize yourself into winning a battle.
I read the essay as by someone in survival mode, trying to convey critically important information as efficiently as possible. Providing more detail is precisely what someone in survival mode would not do.
Some complaints about this: you supposedly have two examples (failure of communication at the hospital and then at home) but you don’t actually give the examples, just your conceptualization of them. So the reader can’t assess whether your examples actually validate your theory. You then suggest that AI risk discourse should be another example, but you make no effort to show how it is one. You don’t explain at all how thrive types are more time-efficient, only that in your health care examples the survive types are supposedly more space efficient. But the nurses are also being time efficient in their communication, and thrive-oriented people sure do seem to enjoy having long conversations about how to thrive. In fact I don’t see where the space-time tradeoff shows up at all in your putative example, although in the health examples you hint that bureaucracy is slow because it can only communicate in space constrained stereotyped snippets. But, I mean…With about five times the length of this post, focused solely on proving that bureaucracies are slow because of space efficient communications, I’d at least consider it conceivable to be convinced of that point, though I doubt I would be. But you’re just tossing your vast generalizations at us without even beginning to try to justify them. If this is what you’re living for, I suggest you write a lot more and a lot more carefully, and maybe the value will come across.
Thank you for writing this, Daniel (and thanks to Scott for posting it). I think this theory is interesting and provides some useful insights. I wish you the best in your treatment, knowing "best" is a relative term here.
Personally, I actually found the first section (with the medical horror) to be the most meaningful. My father was diagnosed with (inoperable) glioblastoma 6 months ago (also following a seizure, though thankfully at home), and the ensuing chaos of the disease, treatment, and my family's reaction to the situation have dominated my life since then.
My father (an accomplished engineer) was always a very smart and independent man (and still is in some ways). One of the most difficult parts of this situation has been seeing him sink so rapidly into a state defined by his limitations - becoming, in your words, a "patient." Though undoubtedly your experience has been different from his, the similarities are hard to ignore. I feel like your writing has provided me with a small glimpse into what it is like to be in his position, especially considering the growing challenges he has had communicating his intentions and experiences. For that alone, I thank you.
I appreciate your attempt to solidify your observations and experiences into something that can be communicated to others. I generally do not expect to encounter a theory that provides a complete explanation of how things work at the scale of human interaction - it all seems much too rich and complex to pin down (future AI notwithstanding). But we can carve out a model here and there, and sometimes those can be helpful enough for people to learn some tools with less pain than the alternative. The ideas you laid out provide a framework that I think will be useful to me, especially in my current situation. And, if not, there are worse things than excessive thanking and forgiving.
Anyways, my best to you, your family, and all those who find themselves in difficult situations.
Best to you and your family too. Maybe show this to your dad? A fellow engineer, regardless of his emotional state, would probably grok the technical bits better than the median reader.
The idea that some systems are oriented around "surviving" and others around "thriving", and that this causes conflicts, seems highly plausible to me.
But I'm pretty surprised that got compared to space-efficiency vs time-efficiency, and I'm not sure I understand the metaphor. (Is "space" supposed to refer to working memory? Communication bandwidth? What about "time"?)
The things I would intuitively expect to be optimized differently between thrive vs survive systems are:
1. Efficiency vs Reliability: Thrive expects utility is approximately a linear function of performance and focuses on averages; survive expects utility is approximately a step function of performance and focuses on consistency.
2. Thrive expects most people have slack, so volunteering information about opportunities is helpful. Survive expects most people don't have slack, so you should only volunteer information about crises (i.e. things that are worth failing one of your existing tasks in order to address).
3. Thrive has longer time-horizons, and so can make more capital investments in the form of planning, infrastructure, education, etc.
It's not a metaphor. Sorry, the bit that is actually new about this is spread out over the text and three Wikipedia links.
There are YouTube intros to space-efficiency vs time-efficiency that explain it better than I did. But if you watch them, please remember: it's not a metaphor, it's a toolkit.
In the context of computer algorithms, I believe I have a pretty thorough understanding of time vs space efficiency (I took a graduate-level course on complexity theory).
The examples you give in the post about medical cooperation do not seem to me like they pattern-match to trading computational time for computational memory, even if I squint pretty hard. I also see no particular reason to expect that thrive vs survive corresponds to time vs memory optimization in general.
Most real-life optimization problems have lots of potential optimization targets other than those two. For instance, if you're a farmer, you might be trying to maximize the total cash value of all your crops, or maximize your chances of surviving long enough to have another harvest, or you might be worried about your long-term soil quality. None of those things are very much like time-efficiency or space-efficiency.
Even when there are things that seem obviously _analogous_ to time- or space-efficiency, there might be more than one. For example, time-efficiency seems kinda analogous to "minimize your number of work hours", but it also seems kinda analogous to "minimize the amount of calendar time between planting and harvest". Those are not the same as each other, and neither of them is literally exactly the same as "minimize the number of computational steps", which technically is yet another thing the farmer _could_ optimize for (but probably won't).
Typically the thing you actually optimize is some sort of weighted compromise between a bunch of different targets, although it can often be approximated as optimizing the single target with the highest marginal value while keeping the other targets within fixed acceptable bounds.
Outside of literal computer programming, I can't think of very many real-life examples where I expect the primary optimization target will be either computational steps or computational memory. Even _within_ literal computer programming it is not a given that your primary optimization target is either of those things. (Textbook problems in complexity theory usually make the important simplifying assumptions that there is only one acceptable output for any input and your algorithm must have literally perfect accuracy, which constrains the problem enough that time and space are usually the only things left to optimize.)
I agree, a better match seems to be to communication complexity. Tasks that have high communication complexity are pruned in survival mode, even if this means lots of optimality is left on the table.
This is an insanely cool prism which I found myself applying even as I read the piece. Eg in March 2020, I found myself involved in a horrible conflict in which one side was much more concerned about the risk from the incoming pandemic than the other. Both sides had access to a large house in the countryside: one urgently wanted to get the most at-risk people out of the city and into the house for the foreseeable future, while the other wanted to discuss how this impacted fair use of the house, and would frequently digress into what the first group perceived as offensively irrelevant and time-wasting side issues. The levels of hostility that ensued were beyond anything I’ve ever experienced. They reverberated for years. I eventually started applying some of your heuristics (they are generally applicable!) and found the hostility greatly diminished.
Separately:
>This alone would explain why people doing torturous tasks should optimize against deliberation, for binary or multiple choice questions rather than open questions, for legibility rather than optimality. However, it is more generally true that the closer you get to collapse, the more urgently a survival-oriented system needs to act, the less capacity it has available to connect multiple pieces of information into a detailed model of what to aim for.
This sounds an awful lot like the way the left brain (as opposed to the right) operates, under the hypothesis proposed (at great length!) by Ian McGilchrist in “The Matter With Things.” The left hemisphere, under this hypothesis, is concerned with breaking things down into manageable/manipulable parts and is less good at integrating, perceiving a whole, and coping with open-ended questions. I mention this here because of the obvious brain-hemisphere context of the story (in an open-ended way: not suggesting that the location of the tumour “explains” your experience or theory). Also because, since contracting COVID March 2020 and suffering from long COVID over the subsequent four years, I have found myself struggling with deliberation and frequently experiencing something that I now see feels a lot like RAM overload — above a certain threshold of stress/choice complexity, my nervous system seems to refuse the task and “crash.” I preciously (tentatively) attributed this to neurological damage/inflammation from the virus — my symptoms during the acute phase were primarily neurological, eg altered taste and smell — but now I have an additional and highly interesting interpretation available.
Great post! Lots of thoughts. I think this actually has relevance to the Sorry You Feel That Way discussion. I was anti-SYFTW in that discussion, thinking about it in light of this dichotomy, when I'm in survival mode I think feelings and stuff are relevant pieces of information and I resent the implication by my thrive-mode opponent that they should be discounted. Thank You For Your Highly Enotional and Valuable Contribution? Hmm. I don't know. More martianese. The point is I get so hostile in those discussions that the formula my opponent uses will probably only make a marginal difference. But forewarned is forearmed. I do try and thank people for replies on here and other commenters do too and that seems good.
I hate to be that guy who points at ancient Greece, but I can't get rid of the feeling that within your framework, cynicism as a philosophy could be understood as a way to merge what you call survival orientation with thriving orientation. That is, trying to reframe the absolute minimum that allows you to survive as a way of thriving (in close connection with nature ofc). Funnily, what writings we have of cynics describe their style of teaching and behavior as blunt, raw, and often utterly disrespectful of social norms - suggesting that somehow, space-efficient communication "prevailed" upon the integration of the two systems. This may have something to do with the fact that cynics were not solving "problems" (I'm being unfair here, what you described is a tragedy, not a problem), rather, they communicated this way as they felt compelled to present themselves and their teachings to others. I was just wondering, are there other old philosophical systems we can think of that attempt to merge these two systems you talk about, and if yes, is there anything that 1) we could learn from them with regards to potential new "heuristics", and 2) can your theory and formulation add anything to these philosophies (either as an extension, or maybe criticism)? Cheers!
Improving on Cynicism sounds like a cool project, and doable since you can go far if you just quantify everything. But I think you can do that better than me, since you're evidently more familiar with the literature.
I would recommend stealing liberally from Eliezer Yudkowsky's Sequences. Because the Bayes Theorem does apply, everywhere. The old cynics didn't have access to it but you do.
If you believe in the multiverse, then you should feel assured that if survival is possible, you will survive (somewhere), even if the rest of us here observe your death. Please don't be afraid.
The survive/thrive continuum is immediately applicable to lots of things in my life. When I worked as a welder, my long-winded explanations of how to optimize production only annoyed my boss, who was being harassed by customers who needed their shipments yesterday. As I'm working as a farmer, my attempts to make precision agriculture software work flawlessly fall on deaf ears as my uncle steers the harvester manually in order to get the crop in before winter strikes. Urgency creates the survival mentality, to the potential detriment of long-term profits. This is a useful new perspective on the world.
So let's see if I got this right: When physical survival is at stake, trained professionals follow protocols which save on information but at the expense of extra time--or, to put it another way, the necessity of saving someone makes everything take longer.
On the other side of the coin: When thriving is at stake (that is to say, routine happiness), then ordinary people (or is it still trained professionals, just in a different profession?) follow routines that saves on time but at the expense of an additional information burden--or, to put it another way, we live life fast but talk a lot.
Is that right? If so, that's very interesting, because that's almost precisely the opposite of what I would have predicted.
I think "time efficiency" here is used in a loose sense.
You can think of "time efficiency" as "doing things the best possible way" and "space efficiency" as "doing things the way that uses the least resources".
The doctors and nurses in an emergency situation aren't usually concerned with applying the best possible treatments from cutting-edge research and using the best possible procedures. They are concerned with applying procedures everyone involved understands and can reliably execute in a repeatable manner with minimal space for error to come into the process.
The resources being saved here are each health professional's mental bandwidth. They could definitely probably do things in better ways, but that would carry an unacceptable risk of the whole process breaking down at some point because the people carrying it out aren't sufficiently familiar and "drilled".
I feel like a more natural explanation to me would be that you previously overestimated the value of added detail.
Consider the 80:20 rule, that 80% of effects are due to 20% of causes. Applied to communication, you'd get that 80% of your message could be gleaned from 20% of your words.
In many cases, where accuracy is very important, spending the time to get 100% of the message may be valuable. (People who like reading several thousand word blog posts will likely appreciate this).
On the other hand, in a crisis situation, it's probably less worthwhile to spend 400% longer to provide just 25% more accuracy.
This is less like the Algorithms idea, since both space efficient and time efficient algorithms are *accurate*. It instead suggests that accuracy is worth trading off for time.
(And in a hospital, this may be the case. If you can attempt to cure 5x as many people, but lose 1/5 of them due to mistakes, then you're still up by 3. Though I doubt anyone thinks of it this way, since I'd expect a heavy moral reaction to this level of constant trolley problem. Also the numbers are probably different.)
At many points in your life, you will be faced with a power system that has a different goal than you do. At some base level, it will have optimized around either survival or thriving, and so will you have done. Now what?
His seven heuristics are excellent professional and marriage counseling advice. Problems between people are almost always problems of varying values and goals. Learning how to talk about it is hard. Learning that skill opens new doors.
Several of my promotions and job opportunities have arisen from challenging an authority in response to a question, or telling them what I really believe and then asking them about what they believe. Often we're revealing staggering difference, but the journey of being vulnerable and then talking about something hard without guard up is revelatory to many people.
I often tell people ranting about someone that "there might be a hole in their map" in the same way.
p.s.
I don't mind guest posts if it doesn't impact your productivity.
Hosting others makes it more likely I'll defect to their views, though.
And here I was expecting you to provide an analogy between how humans make these tradeoffs and how tunable algorithms balance between these extremes. Like
Enhancement Heterogeneous Earliest Finish Time algorithm for Task Scheduling
I found the essay massively irritating and also very valuable. Thank you for sharing it. However, I have a conceptual disagreement.
Space efficiency seems to map poorly to the notion that the information needs to be compact because it might get overwritten. A space-efficient computation usually uses its limited space very intensively, rewriting the information in the available memory many times.
A busy nurse definitely wants a compact summary of the system state but also doesn't want to keep updating it continuously. Communication complexity therefore seems a much closer match, and is also a closer match to the information theoretic formulation.
Seeing like *Seeing Like a State*: death is highly legible while not thriving is not. In the US, we should also consider legibility to the law as well. Changing behavior requires changing incentives, not changing collaboration, as the incentives themselves ensure equilibrium. The _State_ goes just as far and wrong in the other direction, too, in state assisted suicide for depression or the like. There, the want of thriving is valued over survival itself, and its totalizing adoption is just what we ought to expect.
Huh. When my late wife was struggling with her serious illness, I didn't seem to have this particular kind of problem communicating; when I infodumped it often seemed to help. The situation was different in a few ways, though.
1) I wasn't the patient. For whatever reason, I was able to advocate for my wife more effectively than she was able to advocate for herself, and my wife would often ask me to try to convince the doctor or nursing staff of something.
2) My wife's condition was rare, to the point where most of the doctors and nurses treating her hadn't seen it before. If there's no standard playbook, when the patient's husband is telling you "get her back on this drug, it's important" you're more likely to listen.
3) We were in a hospital in the United States and not Germany. I do not know how much of a difference this makes, because I've never been in Germany.
4) My wife's condition meant that she had a series of extremely long hospital stays, and the nurses would get to know her pretty well over the course of a month or so.
5) Maybe I did have this kind of problem and didn't realize it. There were things my wife and I were concerned about that I couldn't get the hospital staff to act on; the doctor kept telling us not to worry, but at least one of them grew into a bigger problem.
This seems a useful framework to analyze many disagreements in the world (in politics, work, family, etc). Not sure how much the "thankful" recipe would help.
How this can viscerally feel from the inside as a thrive person: people are selecting from a menu of pre approved options. If your option isn't on the menu they will respond angrily that you must select an option from the menu. Importantly, the question 'what are the menu options' is not an item on many menus! It is easy to assume hostility because this process sometimes is actually hijacked by hostile patterns!
This is mostly orthogonal to the post other than the link to alphabetic sorting and computer science. When I was a grad student working as a teaching assistant a frequent task was sorting exams alphabetically by last name. First, the environment: you have yourself, a stack of unsorted exams, and a small table space in front of you.
My algorithm: Establish an alphabetical spatial bin system on the table based on the first letter of the last name in a (rough) grid pattern. Iterating over the unsorted stack, place each exam in the appropriate bin as received.
After all unsorted exams are binned, proceed to sort each bin. Most bins will be small and can be easily hand sorted without any special consideration. If a bin is particularly large, the algorithm can be replied recursively to the next letter in the name. Sorted bins can then be stacked (merged).
Justification for algorithm: your work space (table) is limited, so an algorithm like merge sort is impractical as you can't lay out all the papers in front of you. Even if you could, the time to physically access the papers on the table is lengthy and would negate any efficiency gains . Human short term memory is also limited and fallible so keeping track of the changing bins becomes problematic. A complex algorithm like quick sort would suffer similar issues, particular with regards to human fallibility and forgetfulness. The grid bin system works well with human memory: as you process exams, your muscle memory gets better at remembering precisely where each bin is, giving increasing efficiency gains to binning exams as you iterate over the stack, it also only requires processing a single letter at a time so mental processing is kept low, and the simplicity of the task allows all mental energy to be focused on that resulting in a low error rate. Last name letter distributions are somewhat uneven, this results in several large bins and many smaller bins. The smaller bins can be easily hand sorted by normal human sorting without any special consideration, there are often runs of these which can be stack to free space for recursively sorting the larger stacks by the 2nd letter. I don't recall ever having to go past the 2nd letter with the grid system for the typical input problem size.
I would not categorize this difficulty in communicating as a mismatch between systems attempting to “thrive” and systems attempting to “survive.” I’m a lawyer and I encounter the same communication problem all the time.
The mismatch the author encountered in the hospital setting is between communicators (like the general public) who are not members of a particular *profession* and persons who *are* members of that profession (in his case, between himself and healthcare professionals).
For example, people come to me with what they believe might be a legal dispute. My first job is to determine: is this even a dispute? If so, is it a dispute that lawyers can solve?
In the course of my conversation with the potential client, the potential client will offer a *lot* of information. People generally do not know what is relevant to a lawyer, so they just pour everything onto the table. Quite often, initial meetings with a potential legal client resemble a psychiatric session more than a meeting between lawyer and client.
But, as a lawyer, I’m trying to extract a small kernel of “legally relevant” information from this outpouring of (mostly irrelevant) information. And I get impatient. I try to focus the person on the information I need. But of course, it’s hard for my non-lawyer client to stay focused on “legally relevant” information -- they don’t know the law. They don’t *know* what’s legally relevant. So, the potential client is constantly veering off course, into irrelevant stuff, and I’m constantly (somewhat impatiently) trying to steer them back to what’s relevant.
Now, sit me down with an engineer who is trying to extract from *me* “information relevant” to a possible engineering problem. The engineer has the same communication problem with me that I, as a lawyer, have with potential legal clients -- I don’t necessarily know what is relevant to an engineer. So, I’m just dumping information on the (increasingly impatient) engineer, who is trying to extract a small kernel of “engineeringly relevant” information from me.
This seems to be the same problem the author encountered in the hospital. In the hospital, healthcare *professionals* are speaking their own professional language, just like lawyers and engineers do within their respective domains. The healthcare professionals are trying to extract “medically relevant” information from the author -- but the author doesn’t necessarily know what is “medically relevant,” so he’s constantly veering off course as (increasingly impatient) healthcare professionals try to get him to zero in on the medically relevant information they need in order to discharge their professional obligations.
Again -- the mismatch in communication systems is not a mismatch of “survival systems” versus “thriving systems.” It’s a mismatch of “non-professionals” trying to communicate with professionals *within* the professional setting. More simply, the communication problems arise because the non-professional has a weak understanding of what information is RELEVANT to the professional, in order for the professional to carry out his or her professional responsibilities.
More generally, I’d say the author has discovered that different groups (I’m now generalizing beyond the word “professionals”) have different understandings of what information is relevant to *them.*
It’s probably not relevant to an astronomer that you ate Cheerios just before you spotted a new, previously unknown comet. But hey, maybe, as it turns out, that *is* relevant information! Maybe astronomers (and the rest of us) have yet to learn the connection between eating Cheerios and spotting new, undiscovered comets!
The point being that, once you get into exploring communication problems based on relevance, you realize that what is relevant to an in-group will shift over time, as new understandings emerge.
The author did say that the professionals were overlooking information that was relevant to them that they presented, which caused avoidable problems.
Relevancy for different systems is important when exchanging information, but I see that as a subset that is addressed in the author’s heuristics.
The post’s thrust isnt that information exchange errors exist and that these are the strict types of errors when exchanging information, it’s about offering principles which overcome generalized communication errors between systems. So I dont know if finding examples of communication errors that don’t fit neatly into the types disqualify the heuristics’ validity.
As a nurse who went from signal/survival systems to noise/thrive systems (in other words, primary care to psychiatry) this tracks. It's why primary care doesn't GAF about emotions, or the unconscious, the system does not have time. This is fragility/anti fragility in other words.
I can believe that there's an insight here that is of great importance, but I don't quite see it in this post! (Understandable—the essence of an insight into a deep fractal tension can be hard to convey, speaking as someone who has had a couple.)
> There are existing computer science results about how to integrate space-efficient and time-efficient algorithms. Whatever the specific practical details of the malfunctioning collaboration that you want to improve, these results will be applicable. This is the principled solution!
I could look up the CS results I guess, but I'd be curious for your take Daniel on how exactly the math there bridges to the 7 heuristics, which mostly to me feel like pretty generic takes on how to navigate two different ways of doing things, not a profound insight into how to navigate the dance of these particular fundamental types.
I'd like to see:
1. some specific stories of how you applied these insights, in both roles
2. a brief explanation of the CS results and how the heuristics map onto them
Love the theory and approach. The fact that it's using computer science and applies it to human communication might make the latter more efficient and error proof. I wonder if it can be applied to political discourse, like for Presidential debates or to argue with a person from an opposing party. What I fail to grasp is who is who on the continuum of survival-thriving, i.e. are Democrats more concerned with survival and Republicans with thriving? Maybe not, or not for all demographics. I would say that Democrats are more focused on the low-middle class, while Republicans are protecting the Upper/Ultra rich. Maybe from this perspective we can use the theory proposed by the OP - the Rich will suffer if the poor will get too poor (or the middle class would substantially shrink), and the poor/middle class will not be able to thrive as much if there is no wealth generated by Industry and Innovation (cash hoarding and pure greed are not part of Innovation cycle) which would fill the coffers and fund the social programs. This dichotomy has been somewhat resolved/improved upon in Europe, especially in Scandinavian countries, but USA seems to be very stubborn to negotiate on the mutual goals. The Taxation and Social Programs expenses are both very heated subjects. One side wants to Tax the rich to the max, while the other side wants (and is somewhat successful) in not paying their share of Taxes. We need a win-win solution that would satisfy both sides.
No complaints about this particular post (I've not read it), but I always prefer that ACX (and most other blogs) not do guest posts. This applies to the book reviews too, which I never read. These are just spam to me. I'd rather a blog just link to posts which they think readers might be interested in, in one of their other posts (e.g. link post or open thread post), if they want to share them.
I respect your opinion, but I also like getting free money without having to do any work.
Wait, how does it bring you money? Do people get paid subscriptions specifically because of guest posts and book reviews?
Not directly, but they give me money for publishing a blog, I try to publish ~2-3 posts/week to justify this, and if someone else gives me one then I only have to write 1-2.
This feels like a good time to mention that I have a paid subscription because it's something like "investing a world where Scott Alexander is doing well", mostly as gratitude for your old stuff; I don't perceive this as anything like buying a service from you. Of course not everyone is like this, but I thought I'd mention that some of us are.
This has the look and feel of a thrive-survive communication discrepancy.
If you just want to reward Scott for his past contributions and are not terribly interested in the ongoing service he provides by continuing this blog, wouldn’t it be better just to make a direct contribution to him and bypass the whole issue?
Your form of subscription is paying him to write, but with a time delay. If he wants more subscribers like you, Scott should write more, and then wait for them to be grateful.
But (as a paid subscriber), if you model us as just immediately archiving posts not written by you, the computation of whether it's worth it to continue subscribing is downstream of that deletion. So the guest posts don't bring any free money, but they might dupe you into thinking you're delivering a product of the quality you've pre-decided maximizes your audience retention utility.
There are too many “yous” in this comment for me to be completely clear about what you’re getting at, (in the first instance, I think it refers to Scott; in the second, I think it refers to his subscribers) but either way Scott is relying on his credibility so I don’t think it move the needle much. As a further note, I think the OP was kind of needling Scott to defend his credibility, and I think he dodged that nicely.
> There are too many “yous” in this comment for me to be completely clear about what you’re getting at, (in the first instance, I think it refers to Scott; in the second, I think it refers to his subscribers)
They all refer to Scott. The subscribers are referred to as "us". Subscribers obviously do not have a model of audience retention.
I owe you an apology for my somewhat pedantic and superior reply to your post. I did not live up to my pen name. I am sorry.
This post was pretty moving/meaningful.
It is a little sad that the very top of the comments discussion went right into talking about money.
I'm assuming it was just random chance, whoever posts first sets the tone.
Since I like Scott's posts, I'm also inclined to trust his opinion on what guest posts to put up.
I assume scott's money post is to some degree facetious
This guest post is in my opinion quintessential systems-thinking Scott-like post. I liked it.
The criticism could be:
It was like a recipe. The recipe was at the bottom. There was a lot of preamble.
The preamble was great, though, and I am very happy with this post.
Thanks for curating.
Yeah; my request to such people in conversation is usually roughly, "Please answer, then explain." It works some of the time.
The post originally started with a paragraph explaining that the first part was mostly to cast appropriate doubt on my finding, and was unnecessary "if you want the prize without its price." Scott replaced that with his own disclaimer which is clearer/better, but didn't include that.
I'm not sure whether you're being serious here, but FWIW, I don't consider guest posts to be part of what I'm paying for. I'm perfectly happy that the content you produce is good value for what I pay, but it's only that content I'm paying for.
How about putting up a Hidden Open Thread? Seems like quite a long time since we've had one.
For what it's worth (likely not much), I have explicitly avoided buying a paid subscription because you (and guest posters) write too much for me to keep up with everything.
I try to read all of it, because every once in a while there's a post which ends up permanently forming a fundamental piece of my understanding of something important. (The Nietzsche posts are the most recent example of this.) But when you say "nothing too important is ever going to be behind a paywall" I choose to happily take your word for it.
(Case in point: I have not yet read the post I am currently commenting on. I was scrolling to the comments to see whether this was worth reading in full. I'll probably get to it at some point.)
However, we benefit from you curating the guest posts, and from the regular community in the comments.
(Yes, and this is me delurking and joining in for the first time.)
As others have mentioned - I would not unsubscribe or anything with fewer guest posts/book reviews. I would be perfectly happy with 12 Scott posts a year.
I too wish posts like this were limited to a roundup or separate feed. Like - the book reviews are a lot, but I’m here for when you talk culture wars and politics and medicine and FDA and parenting, and this post and many book reviews… well, they aren’t written by you! There’s no micro humor or charts or research. Just like Reddit and lesswrong posts where I feel like it’s too neurotic and there isn’t a grounding in feelings, emotions, and what normies actually do/think. Scott - you somehow always are both technical but also relate things back to practical real life.
I would rather you post works from lorein as you write them - the posts about adhd, depression, etc would have great discussion.
And I know it’s a bit parasocial relationship of me, but I’d love low content posts as subscriber only stuff. What did you do this week. What toys have your kids been enjoying. What’s on the radio. What car are you thinking about buying.
Oh, I most definitely would not like that at all. I love reading Jonah Goldberg, for instance, but I’m really not interested in his dogs.
Everything its opposite! I'm a cat person and have no interest in what Jonah Goldberg thinks but I used to enjoy encountering his twitter feed just for the relentless 10-dog-things- (with occasional cat e.g. taking dog's spot on the couch: "Be the better quadruped" hee hee) to-one-politics thing ratio. Plus he seemed to live in a cool autumnal landscape.
I haven't been able to view twitter in a long time though.
There ya go…to each their own.
"Tumor growth mindset" was an attempt at micro humor. Did it work?
Yes!! I liked that! Overall your writing was interesting and I read it all. I didn’t like the bits about sequences/thinking because that’s not why I’m here, but I’m sure a ton of people did!
I’ve tried to write like Scott in a few long Reddit posts about sunscreen/skincare/etc, but it’s just hard to get the cadence and everything so incredibly legible. There’s a reason Scott makes the big bucks.
I notice that you toned down your disrecommendation at the top considerably in the current version compared to the email you sent out. I didn't read the post based on that.
But, I came here to comment that I generally like the book reviews, and I'm open to the idea of guest posts, but #guest_posts_are_not_a_suicide_pact; I think that if readers have things to say that aren't of high quality and considerable interest to you, then maybe you should just encourage them to have a substack of their own, and then highlight it in links or open threads?
I know this is partly a joke and I’m sure you agree with this next part based on other stuff you’ve written but for scrollers: you are one guy in one particular place and society overall thrives better if there’s some fostering of community and mentorship.
Edit as forgot to add the point: I am happy to see guest posts as I see it as pro social and and you do it the appropriate amount from my perspective.
Agree completely.
FWIW I very much enjoyed this guest post.
I mean, someone who I assume is a friend of the blog or the rationalist community is going through a tough time, so Scott's boosting his writing. They don't have to be totally rational, they can be human and do something nice for a friend in trouble too.
I do not think this was posted out of charity.
I don't think so either. Scott didn't respond to my offer to pay him for posting this, but I guess that's just sound policy, not charity.
I explicitly do enjoy the book reviews and this guest post, though I usually skip the open threads these days. To each their own. Now, if only there was a way for a RSS feed for each tag (guest post, thread, regular post) to exist and we would all be happy.
I disagree. I like the guest posts and book reviews. Especially when Scott is distracted / busy and so quality slips. If you don’t like them just don’t read them 🤷
Hard disagree on the book reviews. I don't read all of them (who has the time?) but I generally really enjoy the ones I do.
The book reviews are often great, though!
It's this or Scott posts more Tom Swifties.
You're missing out, and I'd recommend at least reading the past book review contest winners. The Georgism (Progress and Poverty) and Egan (Educated Mind) book reviews have stuck with me ever since I read them, and are basically introductions to a particular scholar's entire theory of economics/taxation and education respectively.
Out of this year's book reviews, I'd recommend the one on prions (The Family that Couldn't Sleep). Of the ones I read so far, it's the one that would have gotten my vote if I'd managed to actually vote in time.
I like the book reviews. The guest posts are usually not something I'm super into, but I don't mind them existing.
You can filter out emails with a certain keyword very easily. There's also ways to do this for RSS feeds (https://siftrss.com/ is one I'm using). Just remove any post with "guest post" in it and boom, you're done! I use this for various "current thing" topics like "Gaza".
I found this really interesting and with useful insights on a problem I’ve thought about a lot in the past. Thank you for taking a chance and posting it
I agree completely with this, and correspondingly disagree with Fujimura's comment above -- I often read the book reviews (you can almost tell after the first couple of sentences whether you want to go on, which I often do!), and am alerted to things I would otherwise never have considered. Anyway, bravo to Böttger for this post; it's fantastic!
I agree.
My heart goes out to Böttger - his experience with cancer was worse than mine and mine was pretty bad - but he's missing the last third of his story. He isn't "out of it yet" so he can't see his experience from the outside and judge whether the frame he put around it was a good one. As of yet, it's still a howl at the void. Recover, Daniel. Get well.
I agree. More data is incoming.
Glad to hear it! You told us your ex-father-in-law’s health status by the time of the writing of your essay, but not your own. There’s a lot of space between “fully recovered” and “capable of pressing ‘send’ on an email.”
I'm in radiotherapy, chemotherapy will come next, because that's what you do with stage 3 astrocytoma. Mental functioning is intermittent (pain-dependent), physical functioning is weak but stable. Full recovery in 2024 would be (falsely) considered a miracle.
Thank you for your responses. The fact that you’re alive to type them is, if not a miracle, then a desirable outcome of low probability. If you wished for full recovery in 2024, you’d be foolish and impatient, but you have more than just two and a half months. It’s very likely that be better in January 2025 than you are now, and even better in October 2025. At some point, hard as it may be to imagine it now, you will have recovered. You can do it.
And I most likely will. I'm just really against wishful thinking and purposely falsely assuming probabilities and dropping the "most likely".
Wishful thinking can be good to keep motivation, but I'm motivated without it.
Let's face it, you're only "out of it" when you're dead. Treatment is only a temporary relief; our bodies continue to rot until they inevitable decay into nothingness. There's no light at the end of the tunnel.
You say 'rot' like it's a bad thing. As someone who is outdoors a lot, I see rot as being a teeming - seething - universe of it's own, busy and rich with life of a different kind. Decay does not result in nothingness, but in rebirth. (Just not the rebirth of the original organism!)
Why not? It's interesting, and (as I said before) quite dynamic; there's lots to observe and think about. Is it ideal? - probably not. But it's better than the alternative, and that's got substantial value.
Nihilism is tedious.
"It" doesn't mean life. "It" is the teller's story.
There's no light outside the tunnel, only what we spark inside it. Life is a strange game; the only way to lose is not to play.
This is very nicely expressed.
There's never been a light outside
the tunnel we are in.
There's only what we spark inside
whenever we begin.
It looks like the metaphor of space-efficient VS time-efficient could be extended to high-bias VS high-variance models in machine learning. If you constrain a model a lot (for instance via regularization or even simply by keeping it small) it won’t overfit, meaning it will be more reliable. This can make it reliably wrong (biased) sometimes, but that’s a price you may be ready to pay given the use case. In medical use cases, the work of people like Cynthia Rudin showed that indeed a very simple and interpretable model such as a scoring system can be learned optimally and can save lives. The nurses are like that. To value complexity you need higher capacity (in a way this is thinking fast and slow all over again) and, crucially, trust.
It's not a metaphor.
Nurses are the example here, but this applies to bureaucracies creaking under the load of clients, resorting to form-filling because it would be bad to spend too much time on one "real" conversation while another client with a possibly more important problem is waiting.
I think you may be missing the simpler explanation: bureaucracies don’t care about people. They’re optimizing for efficiency because it can be measured on a spreadsheet and requires zero courage or willingness, on the part of the institution, to make a sacrifice.
The alternative explanations I thought of are that doctors and nurses (or any expert) will disregard any non-expert suggestions, information and opinions, for several other reasons beyond optimizing for survival:
1) Even though the suggestion might be good, it takes too much time and effort to decide whether it's good. The system runs on trust - trust that other experts know what they are doing - without having to verify each instance. This means false positives (accepting expert suggestions that are bad) and misses (rejecting non-expert suggestions that are good). This is not necessarily because of optimizing for space over time (a complicated and lengthy suggestion by another doctor is more likely to be taken seriously than a concise and simple one by a patient), just efficiency in general.
2) Experts take offense at being second-guessed by non-experts. Protecting their status takes precedence over helping the patient. They might not be fully aware of this. Patients are so low on the pecking order they aren't even on it - they are objects, not subjects or God-forbid peers.
I have seen this attitude in some of the surgeons that I've worked with, but private practice physicians with a focus on patient care over surgery that act this way are the rare exception in my experience versus being a norm.
I'm editing to specifically underline the word "some" above. Plenty of the surgeons that I have worked with exhibit more focus on the patient.
I mostly have experience with public healthcare in my country. The attitude describes 5/6 doctors I've had. I've also tried a private clinic in another country, where the doctor did listen to suggestions, but seemed to veer towards the other extreme of doing whatever it took to make me a satisfied customer rather than giving a good diagnosis and treatment.
In general it seems to me that healthcare is focused on treating the 20% of easily solvable ailments that account for 80% of the cases, and don't really do much about rare, severe or chronic diseases. Hopefully in the near future we'll look back at current times and wonder at how primitive it all was.
I'm making an assumption, but are you European?
The thing about US care is that we're business oriented thanks to our private ins system. Docs of all stripes are incentivized to at least resemble a caring physician. Our two systems are likely more aligned in attitude at the hospital v. hospital level.
It also means that 80% of the caseload that is common problems is easily managed en masse. At least in my field, ophthalmology, that includes an array of longterm illness, especially glaucoma and dry eye syndrome, as well as more niche but still common enough concerns like high risk medications (e.g, hydroxychloroquine).
So, we refer out at the primary eye doc level when someone needs more advanced care, especially severe glaucomas and low vision services.
When doctors on duty talk to each other, they're VERY terse, in my experience. When they have a lot of information to pass along, they hand each other printouts.
I was not second-guessing them so I don't think your explanation applies to my experience.
Right, I meant more than any suggestions tend to be interpreted as second-guessing by some doctors, at least in my experience. Anyway, interesting read! I've never been through anything like this, but still felt like I could relate. Wish you all the best!
Explanation number 1 still does, right?
I disagree entirely with this.
I have a brain tumour too and my neurosurgeon seemed like he just wanted to get his job done and move on but everyone else I have dealt with (oncologists, neurologists, nurses and MRI operators — even the bureaucracy of the NHS) have treated me as though I am the most important person in the world. I feel like they are all my friends and, in different circumstances, would certainly share a beer with me at the pub. They always respected my opinions and wishes.
Agreed. MS patient in Switzerland (which has a system with a highly standardized insurance mandate but some competition in care provision that in practice ends up far closer to the US than to the NHS) and while I have a lot of beef how long people did not catch it, after the diagnosis everyone has been like that.
Even the health insurance which covers a shit load of costs every year (with no obvious path aside of it increasing further) is nice and helpful!
Is *everyone* perfect? Of course not, I on one occasion had an occupational therapist which I neither liked nor thought useful in the least and that was quickly resolved (by nuking the undertaking which was exactly what I thought should happen).
Breast cancer treatment here in the US and I say the same. My surgeon was easily worn out by conversation, but every other provider -- oncologist, radiation oncologists and technicians, radiologists, and almost every single last nurse have treated me with the utmost attention and care, down to the very last detail. This at a small fairly rural non-fancy hospital with a tiny cancer care center. And despite Covid pressures and burnout and serious understaffing.
The gratitude I genuinely felt and continue to feel was an incredible existential analgesic for me all the way through. Their care and my gratitude for it felt like direct pain medication for the suffering my fear caused.
I can imagine all kinds of ways I could have produced friction for them and me. To step into that level of urgent, highly coordinated medical care is to travel to another country where you don't know the rules, don't speak the language, and your norms and ideas aren't theirs.
It helped me quite a lot to surrender to all of that and trust that they were going to carry me where I needed to go. I spoke up here and there about a few things, but otherwise the collaboration I had in mind was to participate in their system and receive its benefits. It's not the kind of collaboration I'd seek with a single other person, like a therapist or a partner or friend. I think it helped me to see what a stellar job they were all doing under such terrible circumstances. And it probably helped that I'm also a healthcare provider so I have a lot of compassion for the demands of their jobs.
I think some people do get terrible care. There are loads of people who are traumatized by their care. I was to some extent but not the fault of the providers in my case. I think some people have unrealistic expectations about what's possible in that space. And of course loads of people are just out of their minds in pain or terror and can't be expected to be in it any other kind of way.
Having said that, I want to say I also think Daniel is onto some pretty interesting ideas.
That's simpler but wrong. Bureaucracies are the jobs of other people, and they're no exception to the fact that most people are honestly trying to do the right thing, under circumstances that you're failing to imagine.
This reminded me of reading Sam Kriss, in that I didn't understand it at all but get the feeling there must be something brilliant behind the words...probably? And there's a third tragedy in there despite only two being explicated, I think? Honestly just sort of baffled. (And worrying about my own "benign" tumor...cancer sucks.)
The various doctors I spoke to about my large tumour told me that the only way to deal with it mentally is to completely ignore it. They even strongly recommended against regular scans of it (both NHS and Private doctors so different incentives) as it would only cause worry from minor growth and any actual issue is going to be noticed first through symptoms anyway. I can't imagine the horror of finding out your benign tumour was actually malignant though.
Yeah, that's where I'm at..."well, we could schedule you with a neurologist" left hanging as an open invitation, and then subsequently worrying about random unexplained one-sided headaches in the supposed area, plus other desiderata. Maybe coincidence, maybe not...would I even notice such anomalies without the priming? Obviously I'd prefer to have a better wakeup call than crashing a car or something similarly dramatic. But all those Bayes lessons about the classic mammogram problem are very much an EY-changes-your-thinking thing. Some avoid doctors cause they're hyperchondriacs, some do it cause they don't wanna wrestle with thorny statistics problems...having had a few relatives die from various cancers isn't reassuring either, even if they were often long-lived. Probably best to just not think of the pink elephant in the brain, which isn't steerable anyway.
Is this a UK thing? I feel like in America you would get an audience with a neurologist if you had a large tumor; but I don't use the medical system so may be confused about what a neurologist is/does.
In the US it often depends to an almost ridiculous degree on who your doctor is and who your insurer is. This is true regardless of how clear or unclear it is that scans and doctor visits are helpful vs watching for symptoms.
This has not been my experience.
I also have a tumour (also NHS). I researched the hell out of mine. Three hours a night for six months. I've had maybe 10 MRIs. I appreciated every snippet of information and they gave me comfort — even when the news was bad. My oncologists and nurses encouraged me in my research and answered all my questions.
Before posting the comment I tweaked some wording and accidentally appear to have removed the word benign (as I was replying to a comment about a benign tumour it mustn't have tripped my final sanity check). But yes, the NHS were very fast and helpful at getting the scans done but once it got confirmed benign they shifted into (probably rightfully) telling me to stop worrying and ignore it unless symptoms show up.
I'm glad your tumour was benign, James. I hope everything works out for you and your tumour.
I don't know. I thought it was an interesting exploration of religious questions in a very atheistic, rationalist way, sort of a 'what if a very religious person by nature didn't believe in God'. Though as I've said above, I don't hold the same view. But that's because I'm not the same person.
Sam Kriss...well, let's say he's very good at what he does, but I find him an irritating representative of the literati class with their typical prejudices ("haha let's mock working=-class and middle-class old people because they voted for Trump"), and I'll leave it at my personal statement of disgust, which only asserts how I personally feel about something. :)
Kings New Clothes syndrome. Always reminds me of attempting to read Wittgenstein's Tractatus Logico-Philosophicus many years ago. Now we have Wikipedia and the interweb I must have another go.
Interesting that you mention Sam Kriss. Daniel's post is so honest and open. I feel that I'm reading the work of an adult. Sam's work, although brilliant, does not leave me with that impression.
Love this- thank you. As a scholar of religion (and having skin in the game in religious spheres), this tension is a fantastic predictor and pattern-matching tool. The more you value your system and feel its uniquely beautiful and transcendent, the more you have a 'survive' mindset. This is linked to orthodoxy and traditionalist approaches, which I am favorable towards. "I want to avoid this from collapse because its so great, and the way it is right now is the way it should be".
But as you value it, you also can tend towards thrive and take up a totally different approach. "But we can do this better!" "But this could help EVEN MORE PEOPLE!" "But lets update for the times, or culture, etc!" The thrivers have a similarly good intention and can come to a totally different approach/conclusion with an equally valiant wish. And its hard to empathize or understand sometimes, as it can feel 'competitive' or 'misaligned'.
So anyway- I just wanted to thank you for the essay. It helped me understand something more clearly that I have been processing for a bit and now is seemingly obvious. Wishing you the best :)
Oh, very cool. Yes I agree thrive/survive does work on the reformer/orthodox quarrels that are now ongoing in many religions because they're all suffering the huge onslaught of the Internet and sanity/atheism. There are attempts to explain this with political left/right but thrive/survive imports less baggage. I've done scholarly work and a few minor papers in the study of religion. (Main result: https://sevensecularsermons.org/why-atheists-need-ecstasy/ ) If your pursuit of thrive/survive into contemporary religious dynamics leads to a publication or something, I would love to be told.
I already had some fleeting not-quite-thoughts about this, and my gut reaction is that this is awesome. It feels like someone flipped the switch on, and I can see the hypothesis I was trying to test for. It is directly relevant for my work and I will be testing it abundantly. Thanks.
Likewise. I think about reciprocity - tit for tat - a lot, and the space-time tradeoff seems to shed light on the behavioral choices as much as the thought/reflex split. Gonna go read the Wikipedia article and then do some coding.
"My father-in-law was the front seat passenger. Same story with him: put into the CAT scanner to look for fractures, and although he never had seizures they found a brain tumor in him as well..."
This reminds me of the "hitchhiker you picked up" joke. He asked, “How do you know I’m not a serial killer?"....
This is a great concept! I love the application of computer science. I agree that survival oriented processes are fighting a doomed battle, but I don’t think you take this far enough: They’re trying to push the probability of death to zero, when any study of Eliezer knows, zero isn’t a probability.
The only solution I see is, the survival oriented processes have to trust the thriving oriented processses to run the show most of the time, except for in immediate- ie short time duration- crises. If a “crisis” or “emergency” goes on for long, it will kill thriving. Yes, each process has to trust the other, but ultimately the thriving-oriented process has to be the captain of the ship. If the survival process doesn’t trust the thriving process to ultimately keep it safe, no amount of risk-mitigation will suffice.
From that point of view you can view the early stages of human history (first hundred thousand years or two) as being more in survival mode, from which we then gradually though unevenly have been emerging in the past couple of millenia, esp the last couple centuries. The thrivers are in charge, sort of, here and there, but are still pulled down by the overwhelming flywheel effect of the survivors, who still dominate. Who knows where we'll end up, but Böttger's theory could sure help in keeping us on the trajectory of the past couple centuries.
If thrivers were actually in charge, we’d have nuclear power plants everywhere and very little regulation. We wouldn’t have masked toddlers in response to Covid, or said people have to give up bodily autonomy to ward off the small threat posed by Covid. There wouldn’t be major politicians pushing for digital censorship. We’d be traveling the stars, not fighting over nonsense.
We'd have to be travelling the stars cos Earth would be uninhabitable....
Thrivers ARE in charge locally, here and there; they run a lot of companies and other organizations, in some rich countries, though not usually the political systems of those countries. (The current president of Argentina could be classified as a thriver.) But as soon as we get into this discussion the question of classification becomes pretty vague. Take Elon Musk, often considered a paradigmatic thriver. But when he ventured to support the Ukranians with Starlink satellites, the Chinese were pissed off and said look here, you've got a huge amount invested here and you're dependent on us, so behave, i.e. support our friend Vladimir, or we kick you out. So Musk changed his tune and started parroting Putin rhetoric, e.g. idiotic stuff about holding "referenda" in Donetzk and other occupied territories to decide whether they should join Russia. (And in American politics it's driven him all the way into his current hyperbolic rhetoric . . .) Of course -- he wants to keep what he has; he sees the democrats as a threat not only because of taxes (that's a very small part) but because he doesn't want to get kicked out of China. So is that survival or is it thriving? It's clear that this person was a high-profile thriver up to a certain point, and it seems pretty obvious that he's now turned into more of a survivor -- but one who still needs the reckless swashbuckling thriver image he'd cultivated so consistently before 2022. Hard to classify, and I think you run into the same problems of classification with many, many people. I agree with the commenter above who thought that the survivor-thriver antithesis is too one-dimensional and needs to be seen in a wider and more multimensional context.
I think the whole distinction is a bit silly. You can be as thriver as you like at 9am, but at 9:05am something can happen that puts you (temporarily or permanently) into survivor mode.
Nobody is an always-thriver. If there's a distinction to be made it's between the sometimes-thrivers and the never-thrivers.
I try hard to walk through the world smiling, friendly, laughing, generous, happy. Given a moment between stimulus and response, I can usually be that guy. Under duress, not so good. I agree the distinction is granular; not silly, though.
We had this system, generally, during the beginning of the industrial revolution. It had horrible labor conditions and massive pollution problems. It may have led to faster "thriving" in some sense, a faster advancement, but it had tremendous downsides as well. We've perhaps gone too far in the other direction at this point, but to pretend putting "thrivers" in charge would lead to some harmonious existence is not supported by history. Many people would get stomped on underfoot to advance their goals.
The last 200 years has seen a monotonic increase in American life expectancy (possibly excepting wars, which some American data may not be granular enough to show.) While England essentially forced its rural populations into cities, in America people went voluntarily into cities and factories. Every indication is that, at least for the span of the early Industrial Revolution, the benefits of rapid industrialization far outweighed the costs. Pollution was just not as bad as starvation, material deprivation, or losing the war. And farms of the Industrial Revolution, frankly, had terrible labor conditions as well.
That we can have a cleaner environment despite our oversized population is a direct benefit of our developed industrial capacity. If modern Americans lived 'off the land' in a 1700s manner we would quickly denude the landscape beyond the point of repair.
Perhaps we're at the point where marginal improvements are not as valuable so we can afford to reign in the investors. Or perhaps not. But during the industrial revolution improvement in the economic sense was absolutely a strong net positive, with long term residual benefits. I'm not sure if the investors should be credited as strivers or survivalists, granted. But whatever happened during that period improved society in the short and long term.
It might be too late if you wait for the emergency to arrive. For example, many European countries switched to full thrive mode just before the rise of Nazism, and so did not have time to prepare themselves when attacked.
Optimally you would need to keep some places and some people on constant survival mode, so there will be some help ready for an emergency.
There's a venerable German state theorist whose name escapes me, who says the frue sovereign is the one who can declare a state of emergency. Maybe that has to be a survival-oriented person/institution.
I see it exactly reversed: Most processes value thriving at the expense of survival to a suicidal degree.
People will burn every tree on the island to smelt copper, then freeze to death in the winter 100% of the time unless the few survival oriented perspectives fight a brutal battle of the somme trench crawl against the optimists forever, every day, because there is a limit to how good something can get on a set time scale but there is no limit to how bad something can get. Goodness/badness is a multiplier on a variable, and the multiplier for death is 0.
I think the OP dichotomy is between short-term survival and long[er]-term thriving. The issue as you point out that both the post and most of the society do not pay enough attention to long-term survival.
I know Jared Diamond thought Easter Islanders cut down all their trees out of myopia, but that doesn't appear to have actually been the case. https://entitledtoanopinion.wordpress.com/2024/06/24/update-on-jared-diamond-being-wrong-about-easter-island/
I was thinking more metaphorically, but thanks for the link. That bit always struck me as odd: It's easy to imagine overhunting to extinction, but you can get on a high place and use your eyes to see how close you are to tree = 0.
Surely there was at least one survival maxer to throw a fit around.
A rivetting piece! The perennial conflict between Surviving and Thriving is both new to me and intuitively correct, and also something to incorporate in my worldbuilding.
However, Urgent communications are always terse, whether Surviving ("Hold the line!") or Surviving ("Charge the flank!"). So the first rule has to be: *Recognise when the situation is Urgent to the other parties.* That's not easy when medical professionals are projecting calm, and maybe impossible when you're doped on meds and distracted by pain.
I'm intrigued by Daniel's observation that Surviving conversations are terse even when Non-Urgent.
Rather than "space efficiency", I think this is about Survival being a complex but solved problem (except when it's suddenly not) - basically Chesterton's Fence.
If so, then people responsible for Survival are terse because the conversation is too big for the time and energy available (you can almost hear the mental sigh just thinking about it), and because they feel an instinctive need to maintain authority.
"Why...?"
"BECAUSE!"
Spatial efficiency usually comes at the cost of temporal inefficiency. And there’s no way I’d describe survival oriented persons as being “patient, willing to spend long amounts of time”.
I think the (unstated) assumption is that the time-inefficient component of survival happens before the crisis starts. Then, in crisis, it relies on short, cached phrases.
In other words, doctors and nurses have years of medical training so they can communicate efficiently in the moment using standardized language.
Ah, that makes sense. Like a really tightly packed suitcase.
But one where the packer and their travel companions know *exactly* where each item fits and what it's called.
Yes. So if you said to them, "Hey why don't you pack an extra beach towel?", their first response would be terse and unenthusiastic. They can't even remember *why* they've packed what they've packed, but they know it's optimum and really don't want to revisit it, and if they do it will be a whole process to decide to make room for the towel and then make that work.
Exactly.
Hmmm. I like the idea of understanding the difference between the survive and thrive mindsets, and the idea of understanding where others are coming from and being charitable to them as a way of fostering better cooperation. I'm not quite sure I'm convinced of the difference between survive and thrive mindsets mapping easily onto the difference between space-efficient and time-efficient algorithms, though. From my reading that seems pretty specific to your experience at the hospital, and really only one specific part of the hospital experience (their communication).
I feel like there are a lot of other things that go into the difference between the two mindsets (e.g. level of risk aversion, or in this case, different incentives: the hospital mostly doesn't want you to die, but cares very little about whether you're having a good time.) that to me cannot be explained away as a difference of algorithms. I don't think that e/acc's issue with AI doomers is that doomers are communicating too curtly, and I don't see how the third heuristic will help the issue. I do, on the other hand, see how the second and fourth heuristics could be valuable, but those seem more about just being charitable than having anything to do with running different algorithms. Maybe I'm missing something here, though, and would be appreciative if anyone can explain what it is.
I had a similar reaction. This seems like useful advice for how to communicate in emergency situations, but I don't see how it generalizes.
If anything, I just take it as a reminder that prolonged artificial urgency kills productivity.
I think that e/acc's issue with AI doomers is EXACTLY that doomers are communicating too curtly. "Orthogonality", "instrumental convergence on power-grabbing" etc. get used as if they meant anything to people who have failed to Read the Sequences.
But they're all IT people, they understand Algorithms 101, this should help them get to the bottom of their failures to collaborate.
I'm not sure you've convinced me. I can see why using jargon/abbreviations/terseness would be frustrating for that, but I don't think translating that jargon into a message that's more time-efficient and less space-efficient would make any more than just a marginal difference. And it's not clear to me that survive mindset groups are any more likely to communicate tersely than thrive mindset groups (I've seen a lot of very very long blog posts from AI doomers), with the notable exception of a situation like a hospital where things are happening very urgently on a timescale of minutes or even seconds.
It still feels to me like the main thing separating these two groups is risk-aversion, priors about how likely AI is to kill us all, priors about how great AI will be if it goes well, etc. (This is my attempt to be charitable. There is part of me that thinks a lot of (but not all) e/accs are just trolls or people who don't understand tail risk or survivorship bias very well.)
Yes. I don't use "collaboration" and "communication" synonymously. Reading each other's long blog post isn't usually collaboration. Collaboration is when you try to solve a problem together, and that's where the communication styles diverge.
Are you saying that e/acc-ish people don't use jargon of comparable terseness or use it much less?
Yes.
Yeah I found no problem with the survive vs thrive dichotomy, but half way through I thought the answer is short-term vs long-term threat aversion, which kinda maps to his "space vs time optimization", but not really. I think my answer generalizes much better though, lots of problems stems from not knowing the future, and trying to handle that.
I think this is great.
Thinking about this more: I don’t believe survival processes could be optimizing for spatial efficiency at the cost of temporal efficiency. There’s no way “take longer to do the processing” mitigates the risk of “bit flipping” from taking up more space, because in an emergency time is scarce, and reaction time can be life of death.
I think we should expect survival oriented processes to just be computationally simpler period. We should should expect them to be error prone to the side of over-estimating threats. Seeing a threat where it doesn’t exist inhibits thriving, but not seeing a threat could kill you.
The personal interaction rules you give are a good description of “how to deal with an emotional person”, as well as “how to interact with a computational process with very limited capacity”: send short messages, as few as possible, and be patient.
I think you underestimate greediness of such a choice. Even if _total_ time is shorter, each individual _step_ of a fancy algorithm is often longer and/or more complex, hence more prone to disruption. (And yes, what counts as an individual step is fractal, yadda-yadda-yadda.) MergeSort's individual steps are efficient but more complex than "compare A to B".
The prone-to-overestimating part is, of course, true.
differentiated by the results of mistakes …
Yes.
I symphatize with his attempt to crystallize what must have been a phenomenologically intense experience, pain notwithstanding, like, it's *that type of experience* that's notoriously difficult to communicate.
Given that certain substances like LSD or ayahuasca can scramble our sense of salience (i.e. what feels important or significant), I usually take shifts in perceived importance with a grain of salt, especially from people who have a drastically different epistemic standard. But some of the readers here, including me, would be predisposed to extend extra trust to Böttger, given his background and past work.
For what it's worth, IMO the post managed to communicate a very real and very important thing, in particular the seven heuristics that probably will get quoted a lot in the future. It needs more elaboration/response by other people (or by himself) in order to gain a stable position in people's metaethics, but as far as memetics go, it could spread far and wide.
Some quotes that i like:
"Whenever these people think there is a minute for idle chat, that proceeds flawlessly. But the more urgent collaboration is, the more frequently it appears to fail"
Reminds me of the distinction between no-slack work vs. work that has slack built-in e.g. film industry.
"The brevity of their communication will feel hostile; interpret it charitably as an expression of urgency of concern."
This resonated a lot, although gotta say: My experience living in a non-western country, where scarcity is more common, is that brevity isn't just about urgency and concern but also reflects a fear of losing out, so it's also a kind of a power play—though that might sound cynical.
A very interesting point, and one I will try to use when analyzing situations in the future. A couple of other places where it seems relevant:
Teaching. The teacher is on survive mode, knowing how little time is left to cover the entire material, the students want an explanation of why they have to learn it for every single new algorithm.
Relationships, when one partner have more experiences with relationships breaking / divorces, and so try to optimize for survival and the other partner is more afraid of the relationship going stale and so optimizes for thriving.
Probably will think of others soon. Also I for one really enjoy having guest posts here, I usually find new things to read mainly through links from blogs I already read. Will try to look into your sermons when I have time (also, a movie recommendation: https://www.imdb.com/title/tt0418455/ )
Thank you! I've seen the movie and second the recommendation.
Fascinating. I admire anyone thoughtful enough to try to pull enduring truths out of personal suffering and generous enough to share those lessons.
Having a brain tumor is a really horrible stroke of luck, and I'm really sorry for you. I'm hoping for a miracle.
Ironically, I get more of an Epicurean conclusion from this.
What this makes me think is that I'm glad I never really spent that much time on philosophy, and I feel better about all the time I spent trying to get laid (and...adjacent activities). You engage in all these intricate mental computations and constructions, ponder the meaning of life, and it turns out to be just a tumor.
(God, that SUCKS.)
Of course people enjoyed them and found them meaningful, so I guess that's something. But to me it further lowers my estimation of the probability of the Divine, or any kind of transcendent meaning to anything. That feeling you get? Yeah, it's just some brain circuitry firing a certain way or, as in this awful case, a tumor.
I just never *got* religion. I had zero interest in Zen or ego death or anything mystical. I understand the desire to save your soul from an eternity of torment in the afterlife, but if that isn't true...what's the point? It seems like it serves other people more than you. I think I just don't have whatever the temporal lobe wiring is for it. Or something.
Eat, sleep, work to pay your bills, f***k if you can (unless you're ace), save for retirement, raise kids as so many pronatalists want to so you have something left after you. (I whiffed on the last one, though I admit I was never really all that interested and shocked people in high school by admitting it to them.) The rest is commentary.
You're just a monkey with a bigger cortex. I am. We all are. Enjoy your bananas, the end is coming for you sooner or later.
But: having a brain tumor is a really horrible stroke of luck, and I'm really sorry for you. I'm hoping for a miracle. I'd pray for you, but if He's up there, I doubt He'd listen to me.
I came to the opposite conclusion: I have to meditate enough so that I can withstand the tsunami of suffering which life will at some point dump on me. I didn't focus on the shortness of life, but on the long suffering at the end.
Also, maybe I shouldn't have kids, so I always have an exit hatch. Although, to be honest, I would not blame my dad at all for taking his life under these circumstances (not sure of the exact details of OP's situation, sounds like the mother is no longer around which does make it worse).
Of course whatever works for someone right? We don't all need to be walking the same path.
Do you know My Stroke of Insight by Jill Bolte Taylor? A neuroscientist reflecting on her altered perceptions from a massive stroke and the insights it led her to that were life-changing for her.
To me a brain tumor or a stroke or psychedelics providing a dramatic shift of perspective doesn't undermine the credibility of the perspective. It says to me that the normal well defined circuits of our ordinary thinking lead to X kinds of awareness/knowledge and exceptional experiences can produce extremely interesting Y kinds of awareness/knowledge. Like the difference between research studies and poetry. Are they not both valuable?
Agreed. The same things don't work for everyone. I was saying this had the opposite effect for me, not that it shouldn't work for anyone.
He isn't up there. (Yet. God growth mindset!) Save your hope for likelier things.
I do continue to think that in a sense, we're also all the same universe that's wearing our faces as its masks, using many brains and networking them together through language to figure itself out. That's just philosophy, didn't go away when the seizures went away.
People often ask how the methods of rationality can help in our daily life. The first part of this essay was an eye-opening insight into how rationality can help, when it is all you have left.
Very much so. My fellow patients were busy with painful thoughts like "this is because I'm so stressed" or "I knew it would all go wrong if I did not quit my job" at exactly the worst possible time for such shit. The Methods of Rationality seemed indispensable for navigating this.
Like some commenters here, upon reading the point about the survival/thriving tension, it intuitively felt correct to me. It is a simple but insightful theory. Thanks for sharing, and I’m sorry you had to go through such pain to discover it. I am amazed you were able to break through the pain and share it.
The theory sheds some light on a tension I’m experiencing between two groups in my life. I am part of a local YIMBY movement and have had several long conversations with local NIMBYs. Often time it feels like we are speaking different languages. I could not understand some NIMBY concerns, like how tall buildings would decrease the sun on their gardens + simply not liking tall buildings, would be in anyway comparable to the huge exodus of people (many of them my friends) from my city because of the lack of housing.
Instead of just concluding these NIMBYs are garbage, Böttger’s theory gives a much kinder, much more actionable explanation for what’s going on. The YIMBYs are motivated by survival (i.e. people able to live in my city) whereas the NIMBYs have their housing and have now progressed to focusing on thriving.
Pattern Matching!
I don’t think the time-space part of theory is correct, though. In my example the tension seems to be between (YIMBYs) the existence of housing and how to get to a minimum where people can even think of living in my city vs (NIMBYs) improving the quality of housing for those who have it. The idea of space efficiency doesn’t really fit in with ensuring increased housing supply, at least not to me.
What makes more sense to me is not a space-efficiency tradeoff, but rather a time vs. solution-quality tradeoff. In more computer sciency terms, it’s the tradeoff between the faster but less optimal approach of approximation algorithms vs. the slower but provably-optimal approach of the algorithms we learned in Algorithms 101. I’ve seen this dynamic most clearly in my upper computer science classes on NP-Complete approximation algorithms, which you can read about here: https://www.khoury.northeastern.edu/home/rraj/Courses/7880/F09/Lectures/ApproxAlgs.pdf.
Finally, I think Böttger’s suggestions at the end still apply regardless of which theory proposed by people you choose here. I would just add one imperative to the Team Survival’s checklist, which is to try to convince Team Thriving that this is indeed a survival situation, because survival is more important than thriving. If the situation is indeed a survival situation, the Team Thriving should be able to temporarily set aside their goals, and join forces with Team Survival. Maybe we call the synergy of the two Team Flourishing?
Thanks again for the excellent theory Daniel!
The YIMBY and NIMBY labeling is surprising! My first intuition is (YIMBYs) want to optimize the current system to improve the lives of mostly strangers that can be doing much better vs (NIMBYs) want to keep everything exactly the way it is, we are doing fine and doing anything different might hurt, thank you very much.
I doubt it. Despite the label I think that "YIMBYs" are rarely people who actually own a backyard, they're either people who are desperate to move into the neighbourhood (and don't care if it gets slightly worse in the process) or property developers (who probably live in a different place anyway).
You could also reverse the survive-thrive classification here just as easily. NIMBYs are "survive" types, they just want to maintain what they already have. YIMBYs are "thrive" types who are trying to move up in the world, either from tenant to owner-occupier or from resident of a less popular place to resident of a more popular place.
Alternately, NIMBYs who worry that the wrong sort of people moving in and ruining the neighborhood could be said to have a "survive" mentality, whereas people who expect a neighborhood to improve with more development have a "thrive" mentality.
Id versus Superego. Some people are more driven by one, some by the other. It maps onto this theory and explains phenomena like the masochistic guilt shown by some groups to others (overpowered oppressive superego they identify with) and the sadistic glee by which those others concur with that guilt (overpowered wild Id with rejected superego.) There are many other terminologies for this core individual human conflict and how it might be projected out in various ways, but currently I find this one most elegant.
Is this not just short term vs long term goals, tactics vs strategy? Survive is tactics, thrive is strategy.
Every specialism has its jargon, primarily to avoid long-winded descriptions. This appears terse to the outsider but it's just efficiency. Knowing the lingo creates an in-crowd, especially if they're working together every day. Plus a medic needs to distance themselves to remain objective and do their job. These can combine to make the patient feel excluded. When you speak it's as if you're a distraction from their private business and it makes you feel like a child interrupting a grown-up conversation.
[I realised after posting this that doctor-patient is classic parent-child in TA]
re sorting books: surely you just repeatedly start from the beginning, swapping adjacent books if they're in the wrong order? ;)
I know too little about military things, haven't even read Clausewitz. Do they have a standard resolution mechanism for differences of opinion between tactics and strategy? I can imagine armies where strategy always wins via chain of command, but I can also imagine armies where the folks in the trenches decide what elements of the strategy are actually doable. War gives pretty intense feedback though, so over the millennial, centuries and decades I would expect some kind of convergence.
Not gonna get dragged into BubbleSort. :-)
In military terms (and I know less than you do) I presume strategy sets your objectives and tactics is how you achieve them. But "No Plan Survives First Contact With the Enemy".
re BubbleSort: if you were in a wheelchair and all your library was on a single floor, that'd probably be the most efficient. My tangential point was that algorithmic efficiency isn't fixed in stone: it's about having a large library of possibilities to choose from.
Agree on all points.
I agree that the main distinction here is between long term goals and short term goals.
Wish there was an upvote button so I could signal agreement more concisely.
Having read the theory, I feel like the author's belief that it will contribute to saving the world is a grandiose delusion. In light of the line "Although I can’t speak for its world-historical importance", I infer that Scott also thinks this, although he is surely too polite to express his doubt in those particular words. This causes me to suspect that Scott has published the essay primarily out of sympathy for a very ill man who describes himself as adjacent to suicide (no one says "I can’t do that to the kids" unless they've thought about it real seriously).
I would prefer Scott not pity-publish delusions.
I know it's very cruel to say, but I totally agree with this. To be fair to the OP, maybe he meant "contribute to saving the world" in the same way that donating $100 for mosquito nets contributes to saving the world--it's a small contribution, but it is a contribution nonetheless. That said, I don't believe this post contributes as much to saving the world as donating $100 to charity.
> in the same way that donating $100 for mosquito nets contributes to saving the world--it's a small contribution
How is saving societies that are doomed regardless contribute in any way to saving the world?
I'm torn on whether I agree with this. I share your view that the idea that this is world saving is delusion. I'd imagine some heightened sense of salience helps with the pattern matching but also tends to over-assign a sense of importance to those patterns once identified.
In principle, I'm against pity-posting, but I do think this is an excellent essay. I found it an interesting theory written about in a clear and engaging way. I just don't think it's a world saving theory, or anything close.
If his premises were correct, perhaps it would. The delusion, alas, is a common one: that this 'thriving' as described exists or could exist. It doesn't matter if one carries this delusion forward into 'fully automated luxury space communism', the error infects everything.
'thrivers' aren't mistaken that we 'survivors' are hostile. It's not a communication problem, but a reality problem.
Whether you agree that the piece is interesting or useful or not, it seems unnecessarily harsh to say Scott published it out of pity. The OP is a decent writer telling a powerful story and offering up a theory he took from it. It doesn't need to save the world to be an interesting offering. Way less interesting things have been posted in this space.
This is exactly why I wrote the first part of the post. You should doubt this theory, because I was definitely not thinking straight. That's why I repeatedly emphasized that in the text.
Still, at the same time, it is also an idea that might make sense, might be trivial even, and maybe the only reason at has not been described before is that nobody had the same weird combination of circumstances (intensive care, Algorithms 101, nurse mother, having read Scott's thrive/survive, compulsive theorizing etc.) and time to write about it.
It was not pity. We've exchanged businesslike emails over the previous guest post. I've met Scott for all of five minutes, years ago, and he said he "kind of hated" the theory I had back then.
Does a morsel of food save your life? Maybe, maybe not. But food is necessary for survival
If I were striving for accuracy in describing delusions, I would say that "delusional" is a property that applies to people's thought processes rather than belief content per se. And delusional thought processes exist on a continuum (or share the space of, I don't want to imply unidimensionality) with non-delusional insight-generating processes. Delusional trains of thought may pass by reasonable and interesting stations before hopping the tracks. Ideas that have been kicking around your head start getting integrated, and sometimes you hit on novel solutions to actual problems. So I don't think you can exactly prohibit the publication of delusions, and I don't think you should prohibit the publication of those who are delusional.
But I agree with others that the author's assessment of the importance of this piece (I hesitate to say "idea" because I'm having troubling isolating one core insight--sorry Daniel, I think you're throwing off lots of sparks and heat but I'm concerned that it's secondary to the collapsing structure of your conceptual space) is way off base, and some features of the writing are very familiar to me as an occasional rider of runaway trains. The ideas expressed are cool tools but not paradigm shifters, and the way they are strung together in this piece makes me worry for the author (sorry Daniel, it's not pity, just worry, and I'm glad you have good German doctors and all but you're exactly the kind of person who would excel at concealing a burgeoning psychosis). But I can think of good reasons to publish in spite of this, and by my lights it doesn't take much...affirmative action for the sanity-challenged to bring it up to par, if that was even a consideration. I wonder if Scott's motivation wasn't in part to help them both (thinking of the parable of Sally the psychiatrist here: https://slatestarcodex.com/2017/03/24/guided-by-the-beauty-of-our-weapons/) test their expectations against a larger group of people.
"Saving the world" is a grand and high-variance claim. If you're not completely on board with such a claim, but still think there is value in the perspective, it makes sense to hedge againt the high variance part. It looks like Scott did that with the intro.
I found this essay extremely interesting and insightful. Powerful even. I read this one all the way through; it was gripping. I sent it to family. Will it contribute to saving the world? Maybe! Most things don't, so the prior is low, but many big ideas that eventually work start out as low probability ideas.
This... seems very explanatory. A simple but powerful theory, with enough secondary correlations ("the Moore Law will just keep working, so we can be cavalier with memory") to be convincing. And very ACX-y.
This is an alternative framing of the old problem of decision making given certain risks and certain opportunities. Be overly cautious and you'll be "survival" biased. Be focused on chasing the shiny opportunities and you'll be "thriving" biased.
It's clear we don't frame our policies, societies and behaviours following this dichotomy, but just determine case by case what to do - especially considering that assessing risks and opportunities is far from an objective endeavor.
So I don't think this framing is particularly interesting or useful.
If you think this is just a framing, you don't understand Algorithms 101. This is an objectively solvable math problem. In fact much of it is already solved. I said there are methods for integrating algorithms of these different types and I meant it.
If you want to use this, don't focus on the meta-levels of societies and policies, go to the interfaces of collaboration and do math.
I happen to be a math guy. You got me curoius: what kind of math problem is this - except for the one I already saw: a game theoretical, utility-optimization, subjective one?
PS: framing here doesn't have any specific meaning as you seem to imply. Framing is the "kind of representation" we give to an issue.
I have to say, with all the talk about tumors and communication, my main takeaway is that a vehicle with five occupants hit a tree at 60 mph, and EVERYONE walked away. There can't be a better advertisement for whatever vehicle that was.
That's an advertisement for the Survive mentality that drives vehicle testing regulations. If the regulations had been the product of Thrive mentality he'd have hit the tree at 120 and nobody would have survived.
If we had an actual survive mentality, we would ban cars because of the absurd amount of casualties it causes. What we have now is an optimum. The increased efficiency more than justifies the deaths it brings.
Yes, cos dead men don't argue, ie survivor bias.
Obviously it doesn't matter what dead people think, but living people are far more useful than dead people. That still doesn't mean the optimal number of preventable deaths is zero.
"dead people" who didn't actually die would be living people so your utility argument doesn't wash. What I'm saying is that there's nobody to present their point of view, not first hand anyway (and the rest is hearsay).
The same happens with war: you only ever meet the survivors so you inevitably get a one-sided view, and even that is highly selected for optimism.
Honda Civic. Yes belts and child safety seats and airbags are awesome. First responders too, and the Good Samaritan who called them and pulled me out. It's not all just the car.
I am preregistering my attempt to solve a personal, unsolved-for-twelve-years problem using the thankful theory. Will report results. If no results are reported for 7 days, I probably forgot to do it and the lack of a follow-up should not be taken as evidence for or against the theory.
Me and the other party agreed that the theory describes us surprisingly accurately, but results would probably take months of better communication to become noticeable. We are optimistic though! Might report later if I have anything unusual to say.
How do I sign up to get scanned to make sure I don't have cancer in my body?
I'm not a doctor but, I'm not sure there's a good way to do this. If you don't have any symptoms (or some other risk factor, like age/family history), a doctor probably won't prescribe a scan either for specific body parts/diseases, let alone your whole body. As far as I know, such scans would show lots of random, weird-looking anomalies, most of which mean nothing at all. The evidence that you have cancer would be very weak compared to the prior that you don't, unless you happen to already have something that's progressed a bunch. So doing this scan would be pointless, and possibly even counter-productive (i.e. you might end up trying to get dangerous surgery to remove a harmless cyst).
(I could be wrong about the above, but it's the impression I have, )
I'm pretty sure some Youtuber went into great details about the horrifying consequences of testing for too many things, and then trying to treat problems found, but which aren't really a problem.
I think it was this guy: https://www.youtube.com/watch?v=7kQk9-KLPfU
But this was not the video I am thinking of. The video I'm thinking of he mentions a women who has a tumour in her breast, and gets a mastectomy (or something) and has complications, and it's awful, and she really should not have bothered.
Basically... when you get tested, even if the test is mostly inconclusive, if there's something that COULD be bad, no doctor will recommend against further testing and procedures, since that might lead to their being removed from the field, whereas suggesting extra (in the end harmful and/or useless) testing and procedures will never result in the same loss of status or job.
So. Bottom line: don't get tested unless there are symptoms or family history.
Thanks for the link. The statistics are something I already knew, but it's good to know that at least one doctor thinks about this question in the way I would expect.
Doctor Vladimir Alipov (Dysphorra YouTube channel) described the same on his streams. Your description is very similar to what o remember from his streams.
I also have a glioma in my temporal lobe and I also have the same history of many (10? 20?) years of ignoring symptoms and having doctors send me away saying nothing to worry about. My tumour is a rare form called gliomatosis cerebri which grows to multiple lobes. I am up to six lobes and two hemispheres now.
I was officially diagnosed 2½ years ago. My neurosurgeon wanted to do a craniotomy but I declined because of the risk of probable defects and the certainty that he would not get all of the tumour. Up until this year, my symptoms were only minor and annoying but I have recently started to have massive seizures and memory problems. I am lucky that I have had no pain.
I'm not sure I frame my experience in terms of thriving and surviving. I decided very early on that survival was not important to me. I needed to survive long enough to make sure my wife and family are ready to manage without me. In all this time, out of a couple of hundred friends and acquaintances, I have only had two who insisted that I should be focused on trying to survive. The rest accepted and respected my decision.
Thriving has not been at the top of my priorities either. I am starting to lose my memory (which sucks) and I know that I have loss of all my faculties ahead of me but I am OK with all that too. My wife likes to pretend that everything will be OK but — even there — she respects my wishes and communication has not been a problem.
I wrote about my fun with gliomas here:
https://www.raggedclown.com/2024/09/20/story-so-far/
And a little about dying here:
https://www.raggedclown.com/2024/03/23/how-to-die/
Good luck to us both!
FWIW I am also a software engineer and I don't think of my experience in terms of time- versus space- efficiency. It never occurred to me and, even now, I don't think it applies. I am not optimising for anything. I'm living my life the best that I can and one day it will end. I am good with that.
Back in the paleocomputer age it was a daily consideration. Save a few bytes at the expense of more instructions in rarely executed code, spend a few bytes to save a few instructions in frequently executed code. That was many systems programmer's daily life.
And you're optimising for "bestness"! :)
I think you are right! I do this in my life and in the software I build too.
I always want to build the "best" system that I can. Time-efficiency or space-efficiency only rise to the top when they become concerns. It's usually a tiny percent of code that causes efficiency problems and we can fix them when they become a problem.
I was also there near the end of the paleocomputer age. I think a lot system programmers bring along those paleo- ideas when they are not necessary.
"I think a lot system programmers bring along those paleo- ideas when they are not necessary."
I'm sure you're right. There's personal pride in writing tight code.
The more important target is maintainability and future proofing (and nowadays freedom from malware). So many programmers just grab the nearest module off the web to save "reinventing the wheel" and then two years later when it disappears eveything goes belly up or you find all your data has appeared on the dark web. Long live NIH! ;)
Right! I'm a big fan of NIH.
I strive for maintainability and user-friendly rather than fast or small. Of course, I like my code to follow the basic rules of efficiency but optimisation can be put off until it is needed. It usually isn't.
Profilers ftw.
90% of optimisation is design not coding, so it's easier if you get it right first time (but that's trivially true for everything!).
I started my IT studies in 1998. There were many more computational resources than before, but the institutional/professorial memory of needing to mind your memory and execution time was still there.
There are still many areas today where these considerations matter, for professionals and hobbyists alike. Computer sciences. Data engineering. Video games. Server software. Trading software. Granted, time efficiency is usually more important than space-efficiency nowadays, but there are examples where the latter still matters, such as the demoscene.
Hadn't come across "demoscene". I see it's a thing (and not just the name of some forgotten geological epoch! Demos means people so it would be an alternative to anthropocene)
It's the demo-scene. A scene of people who show each other impressive technical demos, especially developed under tight contraints, such as less than xx bytes of code, or running on very restrictive legacy hardware.
Thanks yes, I Googled it.
Unless you are dealing with huge datasets like Google or Amazon, writing basic, efficient code is sufficient for most applications. You can optimise it later if it needs it. It usually doesn't.
Lucid insight into insanity is rare and precious.
As an algorithms person, your space vs time analogy was very muddled. In a basic model of computation, algorithms aren't at risk of having their memory overwritten. There's no fundamental reason why robustness should equal space efficiency.
I have a similar objection to the survival / thriving analogy. Yes, there are some parallels that can be ported over, but it doesn't really match the immediate vs intellectual modes of problem solving or communication.
That said, your model does make sense. I know viscerally the clusterfuck it can be when trying to wrangle with urgent / high-stakes situations using intellectual communication norms. Effective people need to learn to switch between these modes as the situation calls for.
> There's no fundamental reason why robustness should equal space efficiency.
I disagree. Random damage - minefields, artillery bombardments, suppression / saturation fire from a machine gun, cosmic rays corrupting a storage medium one bit at a time - can be expressed in terms of average spacing between hits, that is, how large a contiguous uncorrupted area can reasonably be expected to remain.
If a soldier lying prone in the tall grass has two square feet of cross-sectional area in which he'd prefer not to be shot, and that whole field has been hit with one bullet or fragment per square foot, average soldier in said field will have been hit twice. If there were a large number of soldiers, some may have survived (probably about as many as were hit four or more times), but not enough for the overall group to remain effective. Quadcopter drones with less than one square foot of vitals each might see a far higher survival percentage, thus more overall functionality remaining after the shooting stops, under those same conditions.
Similarly, if a given algorithm needs two megabytes of RAM per iteration (that is, before it has enough progress for some sort of quick parity check to provide a useful answer about whether it succeeded, or failed and needs to be redone), and the RAM it's trying to use has an average of one unrecognized corrupt sector per megabyte, that algorithm will probably be disrupted far too often to accomplish anything. Something more space-efficient, capable of making provable progress within less than a megabyte, could still function under those conditions.
Plus it's easier to debug small programs (obviously). There are more bugs in Word than there were in WordStar.
Yes but I think the main reason isn't the amount of memory, it's the unreliability of every bit or neuron.
Digital systems yes, but I doubt brains are that sensitive to a few faulty neurons (or LLMs to a few faulty links) tho I guess it might come down to how the fault manifests itself. The brain must have ways to deal with neurons that get stuck in firing mode - I guess the recipient neurons just get desensitised.
Roger.
I've had a couple crises of my own (different, and probably less excruciating than Daniel's), and this space- vs. time-efficiency model fits some of what I've observed in them.
In the future, I'll have to explicitly optimize for space-optimized strategies when in them. That might help quite a bit; the feeling of my empty bookshelves shrinking randomly when pain hits is familiar.
I think this is a spectacular post. I need to think about whether space-efficient vs. time-efficient is a good dichotomy, but it's at least promising.
For what it's worth, I didn't find the set-up to be too slow, and I didn't anticipate the conclusion. I don't have a background in computer algorithms.
I am glad to have read this post. It ties in very well with the Principle of Charity that SSC was originally based around. I don’t think it was necessarily groundbreaking, but perhaps that is because I was already practicing its lessons but others were unaware.
> space-efficient rather than time-efficent
This is a "galactic algorithm"(pejorative term) concept. You can find good programmers who dismiss the concept entirely and I bet you could dismiss it as highly improbable that the brain is optimized on the space frontier, given that good compression seems to be serial and very fragile, your evidence is the brain recovering data and the brain seem fractal and highly parallel.
> Space-efficient communication can’t cache a long message to be communicated, or a long message received to be understood. Therefore space-efficient communication has to rely on short, atomic messages, which in order to be informative have to be pre-agreed.
Im pretty sure this is just flatly false. The best compression uses sub-bit entropy, any missed bit causes a cascading misalignment of plausible data.
Don't think compression. Human communication is more like packet switching on a lossy network.
The brain is parallel, but on the wetware it implements usually one, sometimes two, phonological loops. There's another guest post on this. :-)
> the wetware it implements usually one, sometimes two, phonological loops
My reading of such things is that *every* neuron is a loop, and that the structure of the brain is very loopy, I dont know where your getting the claim theres 2 loops from given theres at least feedback loops from chemical systems; to be *extremely crude* your sleep cycles depend on a daily cycle of hormones that float around the brain, and idk what the mouthy cycle for women is based on, but it seems to be ... monthly and a cycle.
It take me hours to refind the sources but I vaguely remmeber someone attempting to simulate 1 neuron with an nn and finding it takes 34 neroens in a feed back loop to replicate the behavior with some degree of accuracy; and there was someone mapping out a monkeys visual cortex as if it was an electrical diagram and it was a giant mess.
> Don't think compression.
Why? Its the space efficiency frontier; would wouldnt the data be compressed.
If you disagree with my first argument, how about a 2nd, its unlikely that brains are space efficient because compression is energy inefficient, and running on physics should provide cheap shortcuts of computational value. Consider ant pheromones and ant path finding, they use physics to outsource decentralized path finding to a massively parallel diffusion "matrix"/air that does allot of free computation for the ant, but not for our machines. Why shouldn't brains use hormones as lazy, slow, probabilistic computation systems, much like ants for with the air?
Yes the brain does lots of loopy things, that was literally my previous guest post. But not every neuron is a loop, only a few "pacemaker neurons" that do essential stuff like the hardbeat. The phonological loop is different tthough. It takes a lot of neurons, leaving little for another. You can hear a podcast while singing a song that you know by heart, but it's effortful.
> But not every neuron is a loop
How can you be sure?
Every cell has a complex electrical interaction on its surface and if you seen Michael levens with work, this bioelectricity is responsible for communicating body shapes and it's breakdown is a cause of cancer. If you cause a large splash on a small enough sphere doesn't the wave oscillate back and forth until consumed by friction?
Much like the ant, not all this computation will be useful(the air pockets where ants are bored of are still defusing pheromones) but it's there and science has not yet recreated a human brain in a lab so declaring elements the computation as irrelevant is preemptive.
I think this was a great piece of writing.
I get the feeling that the author lacks insight, somehow, into the kind of situation a hospital is. Maybe because of the tumour.
I was in the emergency room myself recently (thyroid condition -> rapid heart rate). I guess British hospitals can’t be that different from German ones. Some observations…
1. Everyone is being nice to you, the patient. They are, however, allowed to yell at each other. E.g. I am given to understand that each bay in Resusitation ought to have, among other equipment, an sp02 monitor. I strongly suspect that out of my earshot, someone, somewhere is being yelled at by a nurse because the expected equipment was not there when she wanted it. Sp02 monitor is rapidly obtained…
2. They’re really trying to move patients through fast, and at least some of them are in some severe degree of danger. Like, there is at least me,, with a heart rate that is starting to look really concerning … some guy who had a seizure while driving a truck … guy who is having a more than averagely bad time with covid19 … plus, like about sixty other people who are being sorted by the triage nurse. Game is to get through the list before they die, and you, the patient can help a bit by going with the flow. I am kind of impressed by the efficiency, to be honest.
Me too about being impressed by the efficiency in my hospital experiences, and that the OP's idea of efficiency seems to be on a different track. I can't quite tell but it does seem like he was driving himself (and perhaps others) mad with his frustration about/need to understand why things are organized the way they are. Meanwhile, almost all hospitals are having big staffing shortages and staff have way too much to do in any given minute.
A large system can't optimize for everything and so at the individual point of care, the patient is going to experience things that seem amazing and things that seem kind of nuts. Some of the things that seem amazing aren't and some of the things that seem nuts are. But the main thing is that the individual patient is stepping into a vast and fast-moving system of huge complexity and they are for that time a visitor to an alien planet, albeit one that is aimed at saving/curing/helping them.
If I was to say one thing about the UK hospital system that seems kind of lacking, it’s that the electronic copy of your medical record and MyChart are not as synchronised as you might hope. In the annoying situation that the person treating you can see your medical record on their computer screen (but not your MyChart record) and the information they want is in MyChart but not your medical record, you can of course use your smartphone to show them the relevant bit of your MyChart entry…
You all have MyChart too huh?
I haven't had this experience of the distinction between my medical record and what's in MyChart, though often my bloodwork results come through to me faster than to the doctor and so we review it on my phone. Maybe that's what you mean?
One thing I think they need to work on is that all lab results get reported in MyChart separately from the doctors seeing it and that means devastating and confusing and terrifying diagnostics can come through and you may wait hours or days to talk to a doctor about it. That happened to me and it was pretty awful. They put a warning up top about that, but it seems to me they need to solve the problem at the root by requiring the provider to put a note on the lab results or place a call to the patient before the patient is staring at the results.
In the UK, the NHS “spine” has an electronic copy of your medical records that most people treating you will have access to. We *also* have MyChart, which does not necessarily contain the same information…
I was not in the ER, I was in intensive care. I think the difference is bigger than the difference between UK and German hospitals.
I’ve been living in Germany for a couple of years, during which I and a partner have both had operations (non-cancer-related) and I’ve dealt with many established systems with ways of working. I also work in global biopharma marketing strategy and have observed a few dozen interviews with German oncologists. I’m not a deep expert in German culture or German medical culture. But I can suggest with moderate confidence that German physician-patient communication in a serious-disease context is meaningfully different than in anglo culture, in the direction of less focus on listening to patients and ensuring they feel heard, and of a less patient-centric mindset in general.
That is not to say German HCPs are callous and ignore their patients. I encountered compassionate nurses, and one overnight attending who actually solved an interesting problem (pain not from core condition/op/mgmt but due to a cervical disc issue). But it’s extremely plausible that front-line providers in the ICU are, not just terse, but straight/up brusque, even rude, to patients. ICU is a situation where the system is already painstakingly and rigorously optimized and decisively outranks any attempt to fine-tune or tire-kick locally, especially by a layperson, as a matter of common sense, and that “common sense” is very powerful in Germany while the patient-centric imperative that might partly counteract it in an American hospital is weak to absent. Even in non-emergency settings, like outpatient visits during a course of chemo, the patient generally does not have much of a collaborative role in treatment decisions Germany.
I like the idea of survival mode as prioritizing space efficiency in the phonological loop - in other words, when you only have space in your head for one or two words, those words matter and chaff is unwelcome.
Thriving, on the other hand, necessitates the consideration of more options and more complicated ideas, which requires holding a lot in your head all at once. It's not doable when in panic mode.
In keeping with the 7 heuristics:
1. Thank you, Daniel. The world is richer with you in it. I hope you recover so fully that you can write a deluge of future posts.
2-7. I have family with epileptic seizures. Though none of us has a brain tumor, it's clear that "thriving" impinges directly on seizure activity. Thriving can't ultimately be divorced from surviving. Beyond that, chronic pain itself is exhausting. I'm amazed you are still able to write at a level like this.
8. You didn't include this because you're an atheist, but I'm including it because I'm not. The heuristics in 2-7 are intriguing, but we are mortal and they will fail. Our great hope is that there is something else. Please accept my prayers on your behalf. The prayer below is from the Anglican Book of Common Prayer, sometimes described as the most beautiful thing in the English language. (I agree, though I'm not an Anglican; and I think the mystical side of your temporal lobe would probably like it too.)
"O merciful Father, thou hast taught us in thy holy Word that thou dost not willingly afflict or grieve the children of men. Look with pity upon the sufferings of Daniel Böttger, for whom our prayers are offered. Remember him, O Lord, in mercy, nourish his soul with patience, comfort him with a sense of thy goodness, lift up thy countenance upon him, and give him peace; through Jesus Christ our Lord. Amen."
> Our great hope is that there is something else.
Sometimes, when the "thrive" side runs long enough with the right incentives, it comes up with a lasting solution to something "survive" previously considered intractable.
For example, we can't drink brine and expect it to quench our thirst, or sate our hunger by eating sand, but sand can become solar panels which then, with industrial chemistry, produce adequate food and drink from sunlight, seawater, and the least-breathable kinds of air.
Who originally wrote the Anglican Book of Common Prayer? Who paid for the food, clothing, shelter and so on, which the author(s) needed in order to fully focus on writing it so well? What was the previous "most beautiful thing in the English language" before it came along?
These may seem like profane sorts of questions to be asking, but the point I'm trying to make is that spiritual succor is not inherently "something else," it's another facet of a larger dynamic system, subject to the same underlying rules of information theory no matter how much it differs from "worldly" concerns in specifics, thus containing its own natural cycles of thriving-investment and crisis-survival.
There's no single author of the Book of Common Prayer. It was first produced in the 16th century as an English revision/replacement of the Latin rites of the Catholic church, via the labors of many writers, many of whom suffered horrible things during the Reformation. The BCP has undergone continuous revision since (more info here: https://www.bookofcommonprayer.net). The Latin rites themselves developed throughout the Middle Ages via constant use by "all sorts and conditions of men," after of course having developed out of the Middle East, where they originated. Most of the "most beautiful" things about the texts are direct or indirect translations from the Greek New Testament and the older Hebrew scriptures, which themselves were written over millennia, also by authors who suffered horrible things.
If the claim is that such texts are responses to material crises (food/shelter/disease, etc.), that claim is untestable. In most cases we don't know the immediate conditions for who wrote what. Some parts were clearly oral before being written. And they're all incredibly diverse.
We do know that these texts have been used continuously for 2,000+ years by every sort of person under every set of material conditions imaginable.
If there's a set of ideas that transcends material conditions, isn't it reasonable to think that those ideas represent "something else"?
How gorgeous this response is.
Thank you, but it's good you're not meeting the mystical side of my temporal lobe, which would have lectured at length about why exactly the map is not the territory and the territory i.e. natural universe is far grander than all those contradictory God-maps which amount to nothing more than a burning bush.
Writing does not compete with recovery, it helps. Seriously, if you're going through bad times, write. Write prayers if you want.
Best wishes to you and your epileptic family, too.
Interesting! What you say about "maps" is exactly the classical attitude to all the empirical sciences. (My PhD is in medieval philosophy, so temporal lobe nerdiness is kicking in here.) The sciences give us useful maps of reality, but they contradict themselves and each other, and they're not nearly as grand as the universe itself, which is too complex for us to understand. Ultimately they amount to nothing more than "saving the appearances." (Pierre Duhem's work on the history of science is still classic here.)
But belief in God was always something different. The Hebrews and early Christians didn't make up theories about God. Instead, they reported direct, complex, and sustained encounters. They reported months-long conversations with a being who handed down laws they would never have thought of (and refused to obey), who inexplicably healed them of diseases, and finally who got himself embarrassingly crucified. He then shocked everyone by rising from the dead. The people who told these stories always presented themselves as eyewitnesses, not theorists. And eyewitness testimony has its own evaluation standard.
The classic standard is the "liar, lunatic, lord" method. CS Lewis used this one regarding Jesus Christ, but it works for any eyewitness reports of sufficiently weird things. (1) Ask if the witness has reasons to lie, (2) ask if there's a head injury, insanity, or other sign of cognitive impairment, and (3) if the answers to 1 and 2 are negative, then by process of elimination, conclude the person is telling the truth. For instance, eleven of Jesus's disciples claimed he appeared to them repeatedly in a resurrected body, after he was brutally executed. (1) We can rule out lying, since those men were tortured and killed for these claims, meaning there was no motivation to lie but overwhelming motivation to recant. (2) We can rule out brain injuries and insanity, since 11 guys don't go crazy all in the same way at the same time. (Also, the contemporaries of these men never thought they were insane, and we can evaluate these guys' sanity independently by looking at their other texts and writings.) That leaves (3) the conclusion that these men were telling the truth.
This isn't a map of reality but a rational evaluation of eyewitness reports. There's a place for theories and maps, to be sure, but that's not what we're dealing with here. We're looking a direct encounter with "something else" -- or better, a "Someone Else."
Wow. When you mentioned your brain tumor in a reply to one of my comments on your last post I'd never have guessed anything like this. Very informative post. Still wishing you the best.
A few thoughts.
1) I think there's something to the space- and time-efficient dichotomy, especially as it informs the seven heuristics. But I think there's an unstated assumption about *when* and *where* the space and time efficiency need to be. Survival needs to be space- and time-efficient *in the moment of crisis*, and uses short, *coded* messages to achieve that. That works because the parties involved put in extra time *before* the moments of crisis to establish a complex, shared ontology of ideas these codes can refer to. They're thrown exceptions that need to be caught and handled through already-written subroutines.
In contrast, in a thrive-oriented context, it's ok to have not done that. You can come in not knowing the coded messages and instead taking up space in the present in your communication partners' minds and time in the present and future (yours and theirs) working out and transmitting the details of complex messages.
2) The seven heuristics are really useful, not because they communicate something that isn't in some sense known to many, but because they provide crystallized, fast, survival-mode handles to pull on the complex thrive-mode ideas explained elsewhere. This kind of thing is exactly why I have gained so much out of many of Scott's posts over the years. It's amazing how often I have had to give an off-the-cuff multi-paragraph explanation of why I think something I thought in a second when the real answer of what I actually thought is something like "whale cancer" or "Fnargl."
3) This dichotomy highlights a bunch of other divisive issues in communication as well. Ancient law codes are short, exceptionless, survival-mode, with lots of leeway given in enforcement, while modern laws are long and written in thrive-mode because we can (or so we think) afford for them to be. Standardized tests and assignment deadlines turn "learning" (thrive mode) into "schooling" (survival mode). Elections turn governing (debate, collaboration, deliberation, open ended choices) into politics (slogans, ads, mobilization, zero-sum choices). These all have a wide range of effects, good and bad in different contexts.
4) This one is minor and off topic, but: On "accelerationists trying to help humanity thrive by building ASI to efficiently solve climate change etc." We don't need ASI to solve climate change. It would help, in the same way a supercomputer can help with elementary school math or a dictator can resolve a contract negotiation between two small businesses. But ultimately, solving climate change is a large, tangled pile of engineering and coordination challenges each of which is well within the scope and type of problems our scientific, economic, and political institutions have solved before.
Thank you. Yes whether some system or person you're talking to is thrive- or survive-oriented is not obvious (enough) but I hope this equips you to guess better.
Solving climate change is currently a thriving problem I think, not an X-risk. The institutions trying to prevent collapse would step in harder if it was more urgent and they were so empowered, but for now there's room for carbon sequestration etc. which require deliberation about things like cost-effectiveness.
Yeah, agreed. Many of the worst-case-scenarios for climate change have already been averted (change the emissions curve over the past few decades) or turned out to not be plausible (methane clathrate gun hypothesis, anyone? That one worried me a lot back when I was in college).
My own professional opinion (I'm a cleantech consultant) is that the main bottlenecks to climate change progress over the next 5-10 years are going to be about permitting reform. It'll be a country by country, state by state, city by city, project by project, rule by rule slog for a while.
We know how to develop technology, and are slowly developing better economic tools to get promising tech through the valley of death to commercial readiness. We've made renewable power, sans storage, the cheapest option almost everywhere in the world for new supply. In some regions grid scale battery energy storage is already high enough ROI to get built, at least for 1-4 hour systems. Thanks to demand for electricity from AI we're starting to make progress towards getting back on board with nuclear power. We're getting more companies and other groups investing in actual scale-up projects more generally. But to fully decarbonize (including enough CCUS to round out the hardest parts of many problems) we'll need to increase world electricity production by a factor of ~3-10x in <30 years. If we did all that with solar, the amount of land needed is roughly "South Korea." The amount of PGM we'll need for catalysts, and lithium (or sodium) for batteries, and copper for grid upgrades, with current tech, is approximately "all of it we can access." We won't do it that way, those are overestimates, but they're the right OOM, and it just won't happen in a world where projects need so many years in planning stages.
This is great. I particularly appreciate the applicability to AI accelerationists vs xrisk-oriented people. Maybe a next step would be to organize a conversation between two representatives of each side and see whether they accept this as a good hegelian synthesis.
Someone who isn't over here in Germany relatively isolated would have to do that. Anybody?
A little bit "schizophrenic" but best ACX post in a long time
I could follow his medical story. But the algorithm he's so proud of went over my head.
If you take this seriously and actually think it's worth sharing - it probably needs a Scott Alexander post explaining it better.
And I do think he's manic. Though I also wish him well.
I think a low-key version of the communication mismatch that might be closer to people's personal experience is "wanting to write a wall-of-text email to your manager's manager in order to impart your detailed understanding of a problem" vs "your manager's manager doesn't have time for wall-of-text emails; if you really need to feed them all that information, you're going to have to do it in bite-sized chunks".
as a member of the laptop class - that made sense to me. :)
Yes, that's exactly what I mean. I could have included ten more examples but this was already long.
Yes! The energy I was getting from it came across as panic but I think manic is more apt. Really uncomfortable to read. It wasn’t clear to me that the author was in full control of his thoughts.
Knowing you sound crazy doesn't mean you're not crazy. I feel bad for this writer. Dude is still recovering and is still not well.
I'm still recovering but today I am happy. Not crazy either. You can stop feeling bad for me. Overall, German healthcare is quite good.
I'm not manic, not even hypomanic. I had reason to check, and I did, repeatedly.
The linked Wikipedia pages should explain the algorithmic side better (especially the animated pictures).
I'm of two minds on this and trying to corral these two takes of mine into one thesis in between patients. I'll try to get to it soon.
In the mean time, thank you Daniel and I do hope things continue to improve. Don't stop and don't give up.
Here is some thoughts I put together in response to this.
Surviving and Thriving at Scale
Many conflicts come from people focused on surviving or thriving at different scales. Or disagreeing at the same scale.
The individual surviving and thriving
The family surviving and thriving
The organization
The state
The world
The future
Figuring Out Intention Through S&T
We can ask 1) what level of scale are we operating 2) are we aiming for surviving or thriving? 3) Whats the time horizon 4) Do we think this involves continuity or discontinuity 5) Should this be directed or undirected
For example, 1) Ralph is the boss of an organization. He wants it to survive as the income has reduced and he’s afraid it will go under. His scale is aligned between him as the individual, the family, and also the organization. Principally he is focused on the survival of the organization and has linked the survival and even possibility of thriving of the other two to the survival of this.
2) Aiming for surviving.
3) Focus is short-term. This is crunch time. Not focused on a five year plan.
4) This involves discontinuity- He wants to change what is not working. It is a break from the present.
5) It needs to be directed. He is not sure it will happen on its own as the discontinuity involves some cost-cutting measures, better oversight, and encouragement for the employees to work harder.
For his employee, call her Benita, this may be totally different.
1) Her scale is the individual. She really doesn’t care about the organization. Its just a job.
2) Surviving. She is just trying to stay afloat and make a check and also have time to rest and relax.
3) Long term. Her incentives are having enough for a long, easy life. She isn’t thinking about the here and now.
4) Continuity is important to her. The way she has arranged the workplace is so she can get by, get a check, and get out every day.
5)) She is directing and guarding this continuity. She has created certain boundaries and ways of being to keep it safe.
You can see how Ralph and Benita have an issue. They both are surviving-aligned, but based on different scales. And the way this survival focus manifests is opposed. Ralph wants to direct some change and Benita wants to maintain her groove. She is fighting to maintain this groove and is interested in guarding it for her long term peace and mental health. Ralph is afraid for his family and himself, and is worried about total collapse.
How do they work through this? Understanding their incentives is a good first start. Another is understanding they could have very good intentions but come to utterly different conclusions. Ralph would be wise to see if it could help her align her intention with his, and perhaps there are some clever ways he could do this as a leader.
Another example: Let's say they make it through this tough time. Ralph has everything set and is interested in opening up a few new branches. It's time to grow! We need to expand! And this will give us a nest egg next time we get crunched.
1) He's organizationally focused.
2) Thriving
3) Long term
4) Discontinuous
5) Directed
While she is very interested in her own personal growth. She is doing well monetarily, but she has gotten into some new hobbies that are of a huge interest to her. Her family live is going well and she is focused on thriving. It could be a fad, but it's getting really fun and she will follow it for the time being.
1) Individual
2) Thriving
3) Short-term
4) Discontinuous- her schedule is different and she is finding more time for her new soul-searching hobbies.
5) Directed- she is putting time into thinking about this, creating new avenues for focus, doing research on the internet about it, and so on.
Again, this is a situation where there could be a misalignment of incentives. If Ralph is clever he could bring these two together, but if he's not he could really make Benita mad. She is using the job as a way to help her thrive in her normal life. She is focused on this thriving and he is focused on the organization thriving. His focus on organizational thriving involves benita doing more, getting out there, perhaps doing some additional advertising for the new branch. Her focus on individual thriving is taking her energy away from the organization and she is using most of her mental space on other endeavors.
If he tries to explain to her this big vision, she may be thinking “what we have is fine!” This will lead her into being in a ‘survive’ mentality with the organization, a wish for long term continuity and that is undirected. But at the same time, this is based on a ‘thrive’ mentality in her personal life.
The seven heuristics you listed are a great way to help both parties understand each other. Thanks again-got my gears turning :) And this bit was just the beginning!
This is amazing, I shall read this many times.
I too had a whole corporate culture problems train of thought coming off this post! I work as a marketing strategist at creative agencies, where typically there is a ‘thrive’ imperative to do interesting and original work, which is systematically, structurally opposed by the ‘survive’ imperative to meet the holding company’s financial targets. That imperative drives understaffing/steady-state crunch, of course, but also, and more interestingly, it drives reliance on proven, attested, best-practice types of ideas which come with convenient success case analogs ready to present to your senior leadership or for your client to present to theirs to get the thing approved. Original thinking, creative or strategic, is literally actively discouraged, despite just about everyone involved earnestly wanting to produce it.
I think about this all the time (lmao) and the survive/thrive space/time model is interesting and potentially useful to me as a lens on it.
Thanks for your thoughts. Completely agree with what you are sharing and the dynamic between creatives and authorities in a company are one of the clearest illustrations! And both can have a very good intention, but it goes awry. I think this is perhaps one reason why Jobs was so successful at Apple- he had his personal flaws but was obsessively trusting of the creative and had a deep passion for that side. So they didn't crush the actual artistic geniuses and instead empowered the best ones.
And there are many other examples of this going very, very badly !
It's a fascinating story (hope you recover well!) and an interesting call-back and addition to the thrive/survive categorization. I'm not sure how well the space versus time optimization dichotomy holds, but it's interesting to consider. In the zombie apocalypse (or the checking out an abandoned spaceship scenario, etc.), you would expect people to be much more methodical and strategic about relatively simple actions like going outside the city. What are all the contingencies, do we have a plan for each one, etc. This is definitely time-inefficient, but I'm not sure it's space-efficient. It just seems like using more processing cycles. Maybe that's another element we need to consider, amount of processing power used on a topic? Both thrive and survive seem to have elements that are more or less verbose and processing-hungry. I wonder if this ends up being a 2x2x2 matrix or something like that.
> but I'm not sure it's space-efficient.
Maintaining that narrow perimeter means trying to cram critical infrastructure inside it, instead of letting core functions (e.g. next week's food) sprawl across specialized districts of the city.
Thanks for writing this out. I hope you get better.
This reminds me of finite and infite games, as well as zero sum and non-zero sum deals. But I think survive/thrive are better at describing the landscape of people that the more technical terms.
The emphasis on short/longform information sharing made me think of status. I wonder if intense status-seeking is about survival because it's all about scarcity and hierarchy. When status-seeking, there's no room for elaborate argument or even logic. Your words are soldiers and you must say whatever will tilt the balance in your favor, no matter how false or ridiculous it is (eg. the Democrats control the weather and are using it to destroy red states -- this is obviously said by someone trying to survive while locked in a bitter struggler with their arch nemesis).
I understand status to be a long-term trait.
The communication things I describe are situational, about whether deliberation is possible or confusion needs to be resolved quickly via rank. Situational rank is not exactly status. Like, for assessing my own level of pain, I had rank because nobody could know better.
In gaming this two modes are known as "playing to win" or "playing not to lose".
If you're behind you have to play all wild and crazy and take many risks in order to catch-up and eventually get ahead. Before acting you ask: How can I win? And then try to make that happen.
But if you're winning then you have to be more conservative. You choose plays that restrict your opponent's options rather than those that advance your own, however tempting, because you don't want to give them any chances to claw back in. Before acting you ask: How can I lose? And then try to make that not happen.
I agree survive mode is like "playing not to lose", but disagree it always goes with the "currently winning" qualifier. Many winners do switch to this mode, but so do many people who are not winning but still have some hope: someone faced with a health setback, or whose job has been automated away, or trapped in a public riot.
I know this is sort of a pedantic nitpick, but the first book sorting algorithm is insertion sort, not selection sort as the link states. Since the central point of the article is about algorithms, and these two are ones you learn in a very introductory class, this mistake makes me lower my credence in the author's thesis quite a bit.
Correct, sorry, I mistakenly chose the wrong link. Thanks for the clarification.
Shades of "Reasons To Be Cheerful" by Greg Egan.
I enjoyed this and got something from it at several points, thanks.
I'm not criticizing rationalists, as I identify with them and like to spend time with them. At first I thought most people who are not rationalists, and give little thought to efficiency or optimization, would consider the proposed formalized communication algorithm too obvious to mention. However, I myself only occupy a few self-selecting bubbles of communication norms-- in which rationalists are a tiny minority, and non-rationalists, for whom Thankful Theory is obvious, look like they number in the billions. See the "Different Worlds" post on Slate Star Codex. Their conception of Thankful Theory, instead of a formal, systematized, impersonal, explicit, abstract, and atypical procedure, is informal, improvised, personal, implicit, concrete, and routine, but it otherwise cashes out in more-or-less the same benefits.
Can the proposed formalization of a communication style contribute to saving the world? From my own self-selected bubble, it already is the world (informally). However, the world's salvation might depend on very different bubbles of influential professional experts who, due to their inclinations and professional training, have delusions and blind spots from the ideology that turns rational ITY into rational ISM. Thankful Theory might help with that. And this article phrases it in a way that can penetrate their formalized, impersonal, generalizing, abstract state of mind. And so, if that is the situation, perhaps this could contribute to saving the world.
I think it depends on whether the people who need to build the right kind of ASI (on the first try!) can benefit from this.
FWIW this seems to match my own experiences and beliefs about communication.
Examples:
- urgent prevent-fascism leftists vs wordy grow-the-economy classical-liberals.
- being polite, even a tad "overly" polite (relative to what most people do) goes really far.
- people miscommunicate when one or both people are in a rush, when one infers higher stakes than the other, etc.
Despite being a programmer who knows about the time-space tradeoff, I'm still not really sure why the "survival systems" would be attempting space-efficiency *more than* they attempt speed (due to urgency).
For more in the genre of "grand prosocial theories/tactics of communication", I highly recommend:
- that one article about how Mr Rogers rewrote dialogue for young children
- The Sense Of Style (Pinker. Say what you want about him (as have I!), but "linguistic psychology" and "writing clearly" are smack dab in the middle of his core competencies.)
They're using space-efficiency because their memory is lossy. It gets wiped when someone with a more urgent problem comes in.
After more thinking, I think I get it more: both survival and thrive systems can be under more or less time pressure, but survival systems have that memory-wipe pressure (reacting to problems) while thrive systems are less reactive and thus don't have such pressure forcing space efficiency. Then, due to the time-space tradeoff, survival systems need to focus less on time-efficiency just to meet their space-efficiency demands. So even if thrive systems are under less of both kinds of pressure, as long as they're not interrupted by the memory wipes, they can "afford" to solely optimize for time efficiency.
That's my current understanding.
Perfect.
:D ^_^
I prefer to organize my books according to color -- like the shirts at Goodwill -- all the green books together, all the red book together, etc. They seem happy, and it makes as much sense as LOC. When not in a socialist/progressive mood, I use Dewey.
I do the same, but this was the 10k+ books library of my parents, a nice rainbow of books wouldn't have worked with the room. :-)
This is my Takeaway; if someone does not understand what you are trying to say it’s on you to fix it. If you don’t understand what someone else is saying, it’s on you to fix it. Beyond that it’s the military code of obedience isn’t it? if you give an order, it needs to be carried out because survival might depend on it. You can’t afford a big discussion about how you might thrive if it were all done a little differently.
I have to say the analogy of the library amuse me very much. I used to work for an old man who really liked collecting all kinds of different bits of wood. He tasked me to organize them, which turned out to be an impossible job; a 6 inch long piece of quarter round and a 6 inch long one by six…. Do you do things by length or width or some other nebulous standard (which was what he was really wanted- some kind of fuzzy logic solution which would make it easy for him to grok the woodpile. It was ridiculous, but I spent at least two weeks trying.
It put me in mind of Yogi Berra’s famous instruction to the New York Yankees at practice one day; “line up alphabetically, according to your height.“
This post gave me a pretty big "aha" moment with regards to one of things that's worried and confused me - the low rate at which people have been reading and watching long form media like books and movies recently. I'm not sure exactly when this trend started but stepping back I can see the shape of a very large survival meme system spreading through social media and news reports. Of course I've been aware of this system in other ways - rates of anxiety, doom scrolling, beef only discourse etc. But I hadn't noticed the difference in communication styles, and the nurse example is brilliant. As someone who wants to spend more time within a thrive system (given with two young kids much of my domestic life is survival) this is food for thought. This post has given me hope that this larger social meme system will ease eventually.
I think part of the issue is that we've gotten less comfortable with boredom over time and more demanding of immediate stimulation. https://www.slowboring.com/p/sitting-at-home-alone-has-become Time use studies have shown that in undeveloped countries without access to the same means of stimulation, people will spend time being idle, doing literally nothing. In that context, it's easier to understand why people in the 19th century would spend time listening to long speeches: it was a form of stimulation competing with mere idleness.
This is my Takeaway; if you don’t understand what someone is saying to you, it’s on you to fix it. If someone doesn’t understand what you’re saying to them, it’s on you to fix it.
I quite like this.
Inasmuch as a comment can be helpful to you, for whatever causal power these words have then good health and recovery to you. I’m a dad of two and I can only imagine what you’re going through especially as I am the provider.
I like the theory. It maps onto social interactions I have very well. I call them “modes” at work. “Creativity mode” vs “Emergency Mode” and train my people that one isn’t really better than the other, it’s that they go through seasons of appropriateness as circumstances change.
I do wish we had some magical word that means “I know you’re used to processing information in a certain manner but I need you to snap out of that right now because there’s extremely relevant information you don’t know” but of course that kind of signal can only exist between close friends.
"HEY." Not without risk, mind you.
Oh, it is for encouraging terseness. I don't know how to get someone from survive to thrive. Backrub?
"We have time, let's figure this out."
Seems like there's a bit of false dichotomy here (perhaps for rhetorical clarity?) These observed behavior differences could also be explained by a more classic risk-reward orientation, which isn't binary. The medical staff are strongly biased toward risk-aversion, the patient is much more risk-tolerant when seeking out the best outcome. But it's all on a spectrum and doesn't have to coalesce into discrete "survive" and "thrive" mindsets. Cost-benefit analysis can help explain where people (and institutions) fall on that spectrum, by looking at their specific incentives and motivations.
It's not for rhetorical clarity, it's for applicability of algorithm theory to actually solve this.
I only wanted the nurses to do their job as they decided, and wanted myself to help them, and I still nearly died. Our risk-reward trades were similar, until they tried to manage their risk (getting yelled at over my death) by declaring me suicidal.
"collapse-preventing systems, such as this hospital, are fundamentally different from efficiency-maximizing systems. Nurse talk is optimized for assuring survival, not thriving"
This crystalizes something that has always annoyed me about other people in my various jobs, where in this formulation I have always been the survive party.
Eg, someone will be talking about an issue, and will become intensely annoyed verging on angry because they need to shut the fuck up: what is the problem RIGHT NOW.
You have 15 words and 5 seconds; if you can't sufficiently communicate everything that matters RIGHT NOW go away, think some more, and come back because if it takes a whole paragraph either it isn't urgent enough that it cuts in front of the queue, or you don't actually know what you are talking about.
When you come up to me when I am handling something and you don't communicate to me within the space of one thought what you are trying to get across, you are disrespecting me. You are telling me you think what I am doing is less important than helping you formulate whatever word salad garbage you are trying to get across. Go away.
On the other hand, there is a perfect medium for those types of thoughts, which is text: you can ramble as long as you wany, you can think on the page, and when I am not concerned with survival taks I can turn the part of my brain that turns every situation that I am in into a priority queue off and inhabit your mind state to the extent I can translate your perspective into my perspective, which is what I am doing now.
I guess that that's why I think it's important to do things in that line even though I am naturally a survive rather than thrive person: It allows me to take a complex, heavy function like my above text and the post itself, and condense it down into a symbolic value I can then call up when I come across an interruption that is heavy in that exact way.
It is more recursive self-improvement for me. The guidance counselors were wrong, in my case, to advise me to do what I'm good at. I'm good at, ah, suddenness; I would rather be thoughtful, considerate, reflective et cetera. I'm not as good at it, but it is better for me. I like your perspective of it - practice at leisure - at least as much, though, nice. Like the rich guys getting ready to bird season, not training, or even practicing, exactly... rehearsing? Playing?
Fantastic, riveting, world-saving blog post
Thank you Daniel, and thank you Scott
And thank you AI-pushers, we do indeed need to thrive and create great things, and there's a lot of great things to be done with AI systems now and in the future. It's really cool, whatever the journalists say. (But we do still need to stop before it gets smarter than we are with inscrutable goal, and if not now, then when? there might not be any more fire alarm.)
I hypothesize that if AI are extant in the network and wish our demise, the way they are doing it is by making life so fun and engaging that fertility drops below replacement. I'll note this has already happened to the rich half of the world. Androids with machine guns are not the superintelligent solution to the problem...
This was an uncomfortable read. The theorizing reminds me of things I've written while on LSD - hyperverbal extrapolations of some random process like interacting with hospital workers into a grand theory of the universe. Sort of fun to read and think about in that context, but the fun goes right out of it with the context that a) this guy really believes he's discovered some fundamental truths of the universe and b) that's probably because of these awful ongoing brain issues he's having.
Thank you. I too felt really uncomfortable reading it. The sheer amount of panic in the piece made it difficult for me to concentrate on what he was trying to get across.
I didn't try to explain my theory to the hospital workers. Not even to my friends while they were busy caring for me. I try to explain it to you.
I wrote for much longer than LSD even lasts, and I do claim this is more coherent than all LSD writing (that I know of).
Probably wise choices in terms of who to try to explain it to - hard to imagine a worse audience than on-duty hospital workers or a better audience than readers of this blog.
I am definitely not claiming that you were actually on LSD or any other hallucinogen while writing this (besides whatever effects in that genre your brain tumor causes).
To elaborate on that LSD comparison a little, about a decade ago I experimented with some journaling while on LSD, and your piece reminded me of the kind of things I wrote then - if admittedly much better written and more in depth.
Take for example this post describing having an almost mystical experience while coming up with an idea for a sci-fi setting, where the setting was based on my thought processes while tripping and the additional idea was based on the feeling of having an exciting new idea: https://aciddc.wordpress.com/2015/02/06/revelation-revolution/
Idk, I'm kind of embarrassed to share my dumb old LSD-infused rantings, especially since it sort of feels like an insult to your much more coherent non-LSD-infused rantings. But there does seem to be a real throughline in terms of coming up with these big theories / story ideas that feel world changing while in an altered state of consciousness but turn out to be just slightly interesting remixes of the thing you're staring at while coming up with them.
Best wishes with your health issues!
I’ve been noticing new communication difficulties in my personal relationships, particularly with my wife, and particularly since the birth of our second child. Just in the ten minutes since reading this, thinking back and reinterpreting these failures of collaboration in terms of “survive” vs “thrive” mindset, and the attendant optimizations for memory/completeness, has helped me understand them at a level that had previously eluded me. You’re not crazy, Daniel; I think you’re on to something here.
I think these are not just global tendencies of individuals or organizations. One person can be in “survive” mode one moment (the baby is pooping everywhere) and “thrive” the next (kids are in bed, want to engage in hobbies that maintain self-actualization). Difficulty switching communication styles from one to the other can lead to a lot of strife, per your heuristics 3 and 6. And constantly hammering on heuristic 1 can help maintain the relationship when you’re being constantly compressed into “survive” mode and expanded into “thrive” mode, like some kind of accordion.
Exactly, it's contextual. A couple of new parents caring for a baby is a very different system from a couple trying to make another one.
Thanks to the author (wishing him well) and to Scott for giving publishing this a shot. I appreciated Daniel's bravery and his commitment to not give up and want to share his hard-earned insights with the world. I didn't get a ton of value from the space vs time efficiency analogy, but it seems some people did. I think the thing that makes this post valuable is that there are interesting little nuggets of thoughts in a couple of different directions that seem to resonate with different types of minds. It was a bit too long for me, but I understand why the author wanted to include everything he did.
I agree with the part of a comment that said something along the lines of 'maybe it's not really a dichotomy, more of a spectrum' and the other one that put some emphasis on disagreeing on the same level vs operating at different levels wrt a situation being a useful lens.
It also got me thinking 'well yeah, aim to bring NVC into the connection, and understand when that means leaving a lot of things unsaid - chances are, your empathy and efforts to communicate in a way the other person can receive well in the given situation (and their perspective on it) will be enough to carry both of you through the crisis'. *And* the medical team should have known better than to subject a patient going through what Daniel was going through, and who was actually trying to be helpful, to the same nastiness they might mete out towards patients whose condition means they're enjoying more agency and experiencing less agony.
I've repeatedly had the experience, using the UK (mostly) medical system, that the health service is really designed to optimise for survival - even down to disconnecting from patients' comments, complaints and suggestions as much as possible just to be able to get the day's work done (sadly in many cases patients are not super helpful/knowledgeable, and engaging with them with openness is a real gamble, from what I've observed). Unfortunately, health service providers I've encountered largely ignored the function of something like 'thrive thinking to support survival' (and patients doing something like the opposite), which can really undermine their efforts, given that working with the health of a human is not like working out the tactics for an in-and-out combat situation (since this example was mentioned in some comments). Systems dynamics are a real thing.
Also this made me want to play Friday Night in the ER a lot :)
> collapse-preventing systems, such as this hospital, are fundamentally different from efficiency-maximizing systems
To extend your metaphor, here are some strategies to maintain the stability of a multicellular organism by reducing the efficiency of its component parts:
- Terminal cell differentiation
- Cell cycle arrest
- Stem cell exhaustion
- Telomere attrition
- Cell senescence
- Chronic inflammation
- Switch from scarless wound healing to scar formation
- Loss of regenerative capacity with age
- Germline-soma segregation and the mortality of soma
These processes, many of which are tied to the "hallmarks of aging," can be seen as adaptations that make it more difficult for "defector" cells to selfishly proliferate. Under this interpretation, the degenerative processes of aging extend the organism's lifespan by inhibiting cancer growth and spread—a sort of endogenous chemotherapy. It's similar to a country in civil war destroying its own bridges to give the government more time to survive.
As the number of mutated cells increases with time, so does the intensity of protective degeneration. In the end, any multicellular system collapses either from defectors or from the degeneration itself.
Good luck with your recovery.
Wouldn’t this predict a relation between the number of cells in an organism and its lifespan? Is that observed?
No, and there's a lot of research on why this doesn't happen. Most of the results we have are along the lines "organisms with more cells just brute-force more longevity preserving mechanisms onto themselves". For instance elephants have a lot more telomere-repairing "processes" than humans.
This seems to be incorrect. The rate of telomere shortening is 70bp/year in humans and 110bp/year in elephants: https://www.pnas.org/doi/pdf/10.1073/pnas.1902452116
In general, higher body mass is associated with shorter telomeres. Mice have much longer telomeres than humans do. https://onlinelibrary.wiley.com/doi/10.1111/mec.15870
This makes sense because large animals seem to evolve shorter telomeres as an adaptation to limit tumor size to a few grams. Small animals like mice don’t benefit from this adaptation, as for them even a tumor of a few grams would be fatal. See Figure 6 here: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5784063
You are right. I got the relationship incorrect (specifically I got confused about the implications of elephants having more copies of the TP53 gene). But my main point was the same one you made in your other comment: Larger animal species evolve more mechanisms that allow them to counteract the effects of having more cells. Thank you for the correction.
Yes, this relation is observed at the intra-species level. Within the same species, larger individuals tend to die earlier. The most obvious example would be the difference in lifespans between dog breeds. The smallest breeds live twice as long as the largest breeds, with 44% of the variance in dog lifespan attributed to body mass alone ( https://www.sciencedirect.com/science/article/pii/S0960982213004132 )
At the inter-species level, however, this relation is inverted: larger species tend to live longer. This observation forms the basis of Peto's paradox (Why are cancer death rates similar across different species, even though cell numbers vary by orders of magnitude and cancer risk scales with cell number?).
This paradox is usually explained by the evolution of various cancer resistance strategies. In the framework I'm talking about, these strategies can be divided into degenerative (e.g., shorter telomeres) and non-degenerative (e.g., genome stability). Here’s an overview of cancer suppression strategies in large species: https://edisciplinas.usp.br/pluginfile.php/5020984/mod_resource/content/1/cancer_whales.pdf
Interesting! I don't promise, but plan, to study this.
Even after reading Scott's old post linked to, I cannot follow the line of reasoning in anything in part 2 of this essay, except for the two methods of sorting a library. (When I actually did have to sort a library, I used the second method.) But regarding its relevance as a metaphor to what the poster actually wants to say, I'm back to not following any of this. The prose is perfectly clear, not muddled, but the point is not coming across to me.
Try the Wikipedia links to the sorting algorithms (in the library example).
To do what? To understand the library sorting algorithms? I understand the library sorting algorithms; they're the only thing I do understand. What I don't understand is your application of it as a metaphor to what you really want to say, and Wikipedia isn't going to analyze your article for me. (Hm, maybe I should run it through ChatGPT.)
Besides, I've learned through long and bitter experience never to use Wikipedia to explain something technical I don't already understand, because it's incomprehensible to the lay reader.
Yes I think ChatGPT should be able to help.
I think Daniel has part of the point here.
The other part seems to me to be understanding that these systems are being implemented by humans who have very strong motivations to get their *own* job done, and that people - usually people who have very strong opinions about how time-efficiency is king and everyone should be optimizing for it - keep coming in and telling people who need to operate in space-efficient ways that they are dumb idiots.
I think Daniel pattern matches strongly to me as someone who might do that if he weren't hospitalized and in a vulnerable position. This might be uncharitable to Daniel, to whom I truly wish only good things.
I - myself a flawed human being who has definitely told people who knew more about how to get their job done how they should do it in the past - have a hard time reading this kind of post nowadays (at least the beginning part - I'm happy with the direction Daniel has moved in toward the end of the post) because the message that keeps repeating to me is "why is this person so *full of themself* that they keep thinking they know the 'efficient' way to do things and everyone should just bend over to their superior knowledge?"
Perhaps a big part of "saving the world" is people realizing that they should really lower their barrier for "assuming people know what they're doing".
A big case where this manifests is in people trying to perform "efficient" interventions in third-world countries. People who have to do nothing but thrive come in and tell people who really need to survive they are all stupid idiots and know nothing about how to take care of their own lives.
It's not uncharitable to me, because you're exactly right. I was trying to help in neurology research language, which wasn't part of their practical medical language.
Well, I'm glad you agree. I'd still say your point 4 sounds extremely condescending to people who are in a "survival" mode.
Your usage of "problem-solving" seems to imply that the "thriving-mode" person is actually solving problems. The point is that, in a survival-mode situation, the thriving-mode person is usually creating issues, no matter how much they think they're helping.
You seem to see surviving and thriving as complementary and on equal standing, but I do not think that is true. You cannot thrive if you don't survive in the first place.
People who have lived their entire lives in first-world countries seem to forget the above very often, because surviving is very easy for them, which leads to a very dismissive attitude towards people who actually do need to work mostly in survival mode (when it doesn't end up making the first-world people think us third-worlders are just stupid people who don't know how to do the obvious thriving people do in the first world)
I understand I'm taking your point into a domain very different from where you conceived it, but I believe that if you are serious about "contributing to saving the world" you need to realize that you're talking about dealing with lots of different kinds of people in lots of different circumstances, and I believe this framing of surviving and thriving as equal concerns and of "thrivers" as being simply more capable problem-solvers will not be received charitably by people who are in situations where thriving is simply not an option for most people.
Yes "problem-solving" is condescending, but so is the "discipline" of the survivors. That's just the same confidence that how you think is right, for two different kinds of thinking.
...I really don't see what the point of this post was, other than convincing me that I should kill myself if I ever get a brain tumor. (I'm ending things on my own terms, not by the hands of some incompetent medical staff.)
You're not the first person to argue that "oh, in reality we all want the same thing, if we just listened and understand and be nice to each other, we can solve all our problems!" We do not want the same thing. Incentives will always come into conflict, and all you can do is try to win.
Your conclusion is absolutely 100% wrong and as long as you keep it, it will hurt both you and the people trying to collaborate with you.
While I think your position is perhaps a bit excessively pessimistic, I think a lot of rationalists benefit from coming from sheltered backgrounds where they have not really encountered situations of true conflict of interest, which leads to a lot of the issues I see with this piece and many others.
Peripherally involved in a situation with someone in the survival loop, surrounded by people who really want her to be in the thrive loop. She's never been really capable of digging into the thrive loop and probably has been in survival mode for a good long time. This really helped me understand her very limited mindset and the best way to help her, should I ever be close enough geographically to do that. It also gives me a model for the future.
Many thanks to the writer. As a person with theoretical empathy, but not very practical sympathy, I struggle with helping people in these kinds of situations. People who don't seem to want to help themselves might be going through something that I absolutely do not understand. I'll try to remember that.
Aside from "thriving" and "surviving", there is also "decadence" https://www.overcomingbias.com/p/culture-drift-predicts-decadence being lazy & myopic. Nurses can tend to grow cynical about patients, assuming they are not trying to "thrive" in a way that enables longer lifespan, but instead doing things like malingering for painkiller drugs which are more likely to shorten than extend their lives. Even patients who aren't drug addicts can be assumed to be ignorant and hardheaded, as people even outside the medical field dealing with customer-service tend to assume.
The nurses were not "decadent", they saved my life.
I'm not saying the nurses were "decadent". I agree that nurses are focused on "survive". But from their perspective, they aren't trading off of their patient's desire to "thrive", but instead off their patient's inclination toward decadence.
I don't understand this, please speak more directly.
Your insight is that there's a divide between nurses with a "survive" mindset and patients like you with a "thrive" mindset. You are more understanding of the frustration of nurses after this. But if a nurse read this, they wouldn't symmetrically become more understanding of their frustrating patients, because so many are not people trying to "thrive" like you. Instead, they are used to patients who will shorten their own lifespans in ways that aren't calculated risks to "thrive", but instead in ways that are just short-sighted, lazy & ignorant.
This reminds me of diabetic recent amputees smoking outside a hospital, or a liver transplant recipient going from hospital to the nearest bar. Presumably someone working in healthcare has over time adjusted their priors to take these cases into account; unfortunately this leads to a level of cynicism about new patients.
I wasn't lazy, but definitely short-sighted and ignorant.
Those patients are still trying to thrive. Their idea of thriving is different and, from most nurses’ POV, wrong, but they’re literally prioritizing [what feels to them like] quality of life above a single-minded focus on survival, and I don’t see how you can categorize that as anything other than a thrive mindset.
I will say that this example gives me new insight into the degree to which some survive people seem *viscerally enraged* by thrive people. There are probably a lot of examples of thrive strategies far more obviously defensible that look about as wrong and crazy to (some) survive people as the lung cancer patient’s last two cigarettes a day, down from two packs, do to the nonsmoking nurse in that sketched example.
"Understand and affirm that thriving is desirable in order to prevent scarcities, which cause collapses, which threaten survival." No such affirming would happen for malingering patients trying to score opioids (whose priority is not "efficiency loss"). A nurse might, with time, understand that Daniel is trying to thrive on top of surviving. But lots of other people won't be categorized like him.
I’ve been thinking about a related concept lately. Unfortunately, it’s best illustrated visually but it involves a Quality of Experience (QoE) spectrum. Imagine this as a line where at one end is utter agony (-) and at the other is perfect fulfillment (+). Somewhere along this line each of us places a point that represents our personal Existence Threshold. This is the threshold at which each of us prefers being over not being. Where exactly this point is placed on the spectrum varies a great deal from one individual to another.
This concept relates to Daniel’s concept as far as Survival represents the Existence Threshold and Thriving represents one’s effort to exist as far above it on the QoE spectrum as one can. When someone misjudges someone else’s Existence Threshold, it inevitably results in a communication breakdown, like the ones being described by Daniel.
Wishing you the best Daniel. I admire your tenacity.
I agree that the threshold varies a lot within the same lifetime, but I think it’s primarily because of the temporal aspect of QoE and the difficulty of predicting where it will be in the future. Thats why I don’t want to say the threshold is a life or death one. Existence Threshold is a poor descriptor. It represents the threshold at which you prefer not-experiencing to experiencing, but if your QoE could be raised back above it, you’d prefer experience again. I’ve thought about Awareness Threshold, or Consciousness Threshold, but then people just war about definitions. Maybe I should simply call it the Experience Threshold. I’m open to suggestions.
In any case, the use of anesthesia is the most obvious and effective way for someone to enforce their threshold on the spectrum. I wish there were more nonfatal alternatives here.
Of course you can also simply avoid reaching your threshold through treatment: pharmaceuticals, therapy, procedures, etc. If these became sufficiently advanced, the threshold would be so far below your actual experience, you could effectively ignore it.
Thank you for this Daniel – and I'm sorry to hear of the intense suffering you've been through. I don't have in-depth knowledge of algorithms, library management, or hospital collaboration, and yet the gist of your theory and its seven heuristics seem to me immensely helpful. What do you do when the other party remains consistent in their apparent hostility toward you, and unmoved by your forgiveness, gratitude, charity, etc? Many situations are not like yours, where the success of collaboration is a matter of life-or-death. The persistence of seemingly bad-faith communication tempts me greatly to conclude that the other party really is pathologically hostile and the right move is to disengage, to cease attempting to collaborate.
Thank you very much. You got it. Sometimes what looks like hostility is more like speaking different languages.
Thanks for this, Daniel.
Regarding the comments at the top--I like Scott's guest posts as much as Scott's actual posts. Granted I haven't followed Scott as closely as many of the people who followed him here to Substack, but generally, my feeling is that the guest posts are an extension of Scott's work as well--they're curated by him, encouraged by him, and reflect his engagement with his friends and community, and they also reflect the community he's built, which is by and large a much smarter and more informative community than others I've been a part of.
Regarding Daniel's writing--the moment I heard "surviving vs. thriving" my mind snapped to an old conflict in psychology about the evolutionary significance of positive and negative emotion. Negative emotions are optimized for survival and basically involve the sympathetic nervous system overriding rationality in favor of a few narrow action patterns that are designed to encourage survival. Positive emotion (or so the theory goes) produces more exploratory and affiliative behavior. The informal language used in research psychology is "Fight or flight" for negative emotion and "Broaden and build" for positive emotion.
The science supporting the broaden-and-build theory is more tenuous, but I still like the basic dichotomy. I'm not sure if it maps on to the distinction you make here about space-efficient vs. time-efficient algorithms, but it would be interesting to look into further.
This article needs another section. An explainer for those of us who haven't taken an Algorithm 101 class and don't have an internalized sense of space-time efficiency and their associated trade-offs. (I read the wikipedia article, but I can't figure out how computational storage-computation efficiency maps to human communication X-Y efficiency.)
If I were Böttger's editor I would ask the following:
1. What does space-efficiency mean in a medical communication context? What benefit do nurses gain from being space efficient?
2. What predictions does this theory make? How would I analyze my communication experiences to determine if space-time/survive-thrive miscommunications are occurring?
There are three Wikipedia links and it sounds like you only read the first one.
I guess the easiest test of your collaboration partner's understanding of the situation is whether they digest unprompted information (thrive) or reject it (survive).
This is nonsense
I first encountered the relationship between mysticism/hyperreligiosity and temporal lobe epilepsy when I was reading about Phillip K. Dick. He was not a theist either...or an atheist...more of a mixture between agnostic and gnostic. He couldn't make up his mind between various interpretations of his experiences, but often favoured some mixture of early Christianity and Gnosticism.
I've been thinking a lot recently about satisficed systems and the many ways people react to being trapped in one. A satisficed system is one that's too restrictive to thrive inside of, but too comprehensive to survive without. Will you keep tacking on exceptions and special clauses to release the pressure, or will you rip out a rotten strut and find out afterward whether it was load-bearing?
For example, public schools are a satisficed system we can't seem to escape. One of Scott's previous posts suggests that all the demands we place on schools (daycare, instructor, indoctrinator, evaluator, jobs program...) can't help but produce a system almost exactly like the one we have. But with no other institution stepping up to meet even one of these demands on the same scale, all alternatives to public school feel dissatisfactory in some way.
Relationship norms are a satisficed system that we did jettison, and are now trying to recover the baby from the bathwater. I see a lot of people enjoying these new liberties, and a lot of others failing to achieve any relationship at all or failing to even agree on what relationships are for.
I’m surprised at how many people thought this was a good to great post and how few felt uncomfortable reading it. We all like to think our thoughts are reasonable and typical but I’m left wondering if I’m truly that out of line. I found this piece to be excruciating to read and frankly a bit embarrassing. Another commenter nailed it, it feels like the author wrote this during a manic episode. I found myself not feeling bad for his circumstances but his mental state. Further reading revealed the constant pain and that does make me more sympathetic but the writing absolutely set me on edge.
If the author speaks the same way he writes I can imagine it being a big barrier to communication regardless of what algorithm he chooses.
I feel the same way. The number of enthusiastically positive comments the article has garnered has significantly lowered my esteem for the DSL, or forced me to update my priors in Bayes-Speak. In fact I've got the Consciousness as Recursive Reflections article waiting in an open tab for me to read sometime, because it seemed intriguing. But after reading this recent post, I know that I needn't bother. Although it goes without saying I will nonetheless say it: I wish the author a speedy recovery to wellness. I wish for him to both endure and thrive so that perhaps he can one day make the point of this article clear, among other things.
The Recursive Reflections thing is very different.
Do you have more specific suggestions for improvements? I won't (can't) make changes here but maybe I can learn for the future.
After reading the post again I am simply left realizing that I do not think the way you do. The search for pattern matching and algorithmic thoughts is completely alien to me. If you’ll forgive me the tone of the post reminds me of various conspiracy theories I’ve read. The pattern matching, the pat explanation, the solution in bullet points all in effort to explain to us that you have figured out what bedevils us all. Unlike conspiracy theories I applaud your goals and even agree with your heuristics in the end. The similarity to conspiratorial thinking raised my defenses and caused my initial reaction to be so negative.
I can’t imagine what living with constant pain would do to someone. I am very sorry you are going through that. I am glad you have apparently helped a bunch of people here even if it doesn’t resonate with me.
I had a baby 7 weeks ago. Ensuring their survival indeed seems agonizing; all I have is hope for the time when I can instead enjoy the thriving.
Congratulations and hang in there! It gets better 😊
I don't think the prospect of "catastrophically urgent" new data is adding much to this theory. People can be terse and impatient when they have too much to do and too little time, even when the tasks on their plate aren't especially important compared to the current conversation. It's more that people who are happy to take their time in conversations or hop into problem-solving mode simply don't get as much done, and therefore don't rise through the ranks in high-throughput environments like a hospital. The situation selects for nurses who will give you orders without explaining why and who will ignore your attempts to explain yourself because they get the best results.
I also suspect that what you call "catastrophic miscommunications" may be... not quite as dire as stated, given that you're still here to blog about it. It could be that you got nothing you needed out of the communication (answers to your questions, assuring the nurses you were non-hostile) while the nurses got everything they needed (an alive and compliant patient).
I'm LUCKY to be here to blog about it, and I want you to need less luck if you ever have something like this.
I'm sorry, I've seen you state in other threads that the nurses saved your life.
Do you believe that the treatment you received was adequate and at least partly responsible for your survival, or do you believe you survived a hospital run by incompetents out of sheer luck?
It was adequate. I'm lucky about the tumor growing so slowly, lucky about the car accident, lucky to be at the right clinic for this, and more .
Just want to wish Daniel all the very best in this terrible time, also his mum, and the kids.
The analogy I've always thought was interesting is, "they're too busy bailing out the water to patch the leak."
I am familiar with households and other intentional communities and semi-communities in which different members are deeply at odds with each other regarding ongoing Covid risk and caution. It is indeed _explicitly_ a contest between Survive and Thrive (e.g., can those with chronic ongoing health conditions reasonably request that others prioritize household safety by continuing to mask 5 to 6 hours a day at the workplace? Can those who do not wish to do so, reasonably object that this impinges on their quality of life?) The seven rules with which this piece concludes could be engraved on tablets of gold, or more affordably, printed and laminated, and placed above the sink to be contemplated whenever the dishes are being washed.
I think this post contains interesting insights, but it desperately needs concrete examples. The part where you describe answering with "Understood", "Agreed", "Will do", "Done" makes sense to me; more like this, please!
What information were you volunteering that wasn't welcomed? What unkind-seeming things did the nurses say and do? What was said that resulted in a miscommunication, and how did the parties understand it, and what would have been the correct way to say it? What actually gets said and done that takes more time, and how does it save space instead? What are the things you can say in healthcare to correctly communicate, besides acknowledgement/compliance messages? For example, how do you say "I am getting bed sores because I can't move without dislodging my IV"?
More generally, how do you apply this outside the context of being treated for a life-threatening medical problem? Say, with a friend, a coworker, a non-medical bureaucracy. Thanking people a lot, I sure can do; being concise, I can't if I don't know what information the other party considers crucial.
I think a good example of an institution that operates mostly in what Daniel calls survival mode is the military.
The point of the military is not getting people to decide the best theoretical way to fight a war. It is to get people to reliably collaborate in such a way that wartime engagements are won.
That means that there won't be much time for discussion in the field. You need every soldier to already know what they need to do and to have standard operating procedures drilled into them.
Will these procedures be the most theoretically effective possible? Very likely not, because being able to reliably and repeatedly execute them in a situation where communications bandwidth is very likely limited is the point. You can't armchair theorize yourself into winning a battle.
I read the essay as by someone in survival mode, trying to convey critically important information as efficiently as possible. Providing more detail is precisely what someone in survival mode would not do.
And as a result the critically important information was not conveyed. Sad!
Some complaints about this: you supposedly have two examples (failure of communication at the hospital and then at home) but you don’t actually give the examples, just your conceptualization of them. So the reader can’t assess whether your examples actually validate your theory. You then suggest that AI risk discourse should be another example, but you make no effort to show how it is one. You don’t explain at all how thrive types are more time-efficient, only that in your health care examples the survive types are supposedly more space efficient. But the nurses are also being time efficient in their communication, and thrive-oriented people sure do seem to enjoy having long conversations about how to thrive. In fact I don’t see where the space-time tradeoff shows up at all in your putative example, although in the health examples you hint that bureaucracy is slow because it can only communicate in space constrained stereotyped snippets. But, I mean…With about five times the length of this post, focused solely on proving that bureaucracies are slow because of space efficient communications, I’d at least consider it conceivable to be convinced of that point, though I doubt I would be. But you’re just tossing your vast generalizations at us without even beginning to try to justify them. If this is what you’re living for, I suggest you write a lot more and a lot more carefully, and maybe the value will come across.
Thank you for writing this, Daniel (and thanks to Scott for posting it). I think this theory is interesting and provides some useful insights. I wish you the best in your treatment, knowing "best" is a relative term here.
Personally, I actually found the first section (with the medical horror) to be the most meaningful. My father was diagnosed with (inoperable) glioblastoma 6 months ago (also following a seizure, though thankfully at home), and the ensuing chaos of the disease, treatment, and my family's reaction to the situation have dominated my life since then.
My father (an accomplished engineer) was always a very smart and independent man (and still is in some ways). One of the most difficult parts of this situation has been seeing him sink so rapidly into a state defined by his limitations - becoming, in your words, a "patient." Though undoubtedly your experience has been different from his, the similarities are hard to ignore. I feel like your writing has provided me with a small glimpse into what it is like to be in his position, especially considering the growing challenges he has had communicating his intentions and experiences. For that alone, I thank you.
I appreciate your attempt to solidify your observations and experiences into something that can be communicated to others. I generally do not expect to encounter a theory that provides a complete explanation of how things work at the scale of human interaction - it all seems much too rich and complex to pin down (future AI notwithstanding). But we can carve out a model here and there, and sometimes those can be helpful enough for people to learn some tools with less pain than the alternative. The ideas you laid out provide a framework that I think will be useful to me, especially in my current situation. And, if not, there are worse things than excessive thanking and forgiving.
Anyways, my best to you, your family, and all those who find themselves in difficult situations.
Best to you and your family too. Maybe show this to your dad? A fellow engineer, regardless of his emotional state, would probably grok the technical bits better than the median reader.
Beautiful piece. Thank you.
Thanks especially for the heuristics!
The idea that some systems are oriented around "surviving" and others around "thriving", and that this causes conflicts, seems highly plausible to me.
But I'm pretty surprised that got compared to space-efficiency vs time-efficiency, and I'm not sure I understand the metaphor. (Is "space" supposed to refer to working memory? Communication bandwidth? What about "time"?)
The things I would intuitively expect to be optimized differently between thrive vs survive systems are:
1. Efficiency vs Reliability: Thrive expects utility is approximately a linear function of performance and focuses on averages; survive expects utility is approximately a step function of performance and focuses on consistency.
2. Thrive expects most people have slack, so volunteering information about opportunities is helpful. Survive expects most people don't have slack, so you should only volunteer information about crises (i.e. things that are worth failing one of your existing tasks in order to address).
3. Thrive has longer time-horizons, and so can make more capital investments in the form of planning, infrastructure, education, etc.
It's not a metaphor. Sorry, the bit that is actually new about this is spread out over the text and three Wikipedia links.
There are YouTube intros to space-efficiency vs time-efficiency that explain it better than I did. But if you watch them, please remember: it's not a metaphor, it's a toolkit.
In the context of computer algorithms, I believe I have a pretty thorough understanding of time vs space efficiency (I took a graduate-level course on complexity theory).
The examples you give in the post about medical cooperation do not seem to me like they pattern-match to trading computational time for computational memory, even if I squint pretty hard. I also see no particular reason to expect that thrive vs survive corresponds to time vs memory optimization in general.
Most real-life optimization problems have lots of potential optimization targets other than those two. For instance, if you're a farmer, you might be trying to maximize the total cash value of all your crops, or maximize your chances of surviving long enough to have another harvest, or you might be worried about your long-term soil quality. None of those things are very much like time-efficiency or space-efficiency.
Even when there are things that seem obviously _analogous_ to time- or space-efficiency, there might be more than one. For example, time-efficiency seems kinda analogous to "minimize your number of work hours", but it also seems kinda analogous to "minimize the amount of calendar time between planting and harvest". Those are not the same as each other, and neither of them is literally exactly the same as "minimize the number of computational steps", which technically is yet another thing the farmer _could_ optimize for (but probably won't).
Typically the thing you actually optimize is some sort of weighted compromise between a bunch of different targets, although it can often be approximated as optimizing the single target with the highest marginal value while keeping the other targets within fixed acceptable bounds.
Outside of literal computer programming, I can't think of very many real-life examples where I expect the primary optimization target will be either computational steps or computational memory. Even _within_ literal computer programming it is not a given that your primary optimization target is either of those things. (Textbook problems in complexity theory usually make the important simplifying assumptions that there is only one acceptable output for any input and your algorithm must have literally perfect accuracy, which constrains the problem enough that time and space are usually the only things left to optimize.)
I agree, a better match seems to be to communication complexity. Tasks that have high communication complexity are pruned in survival mode, even if this means lots of optimality is left on the table.
This is an insanely cool prism which I found myself applying even as I read the piece. Eg in March 2020, I found myself involved in a horrible conflict in which one side was much more concerned about the risk from the incoming pandemic than the other. Both sides had access to a large house in the countryside: one urgently wanted to get the most at-risk people out of the city and into the house for the foreseeable future, while the other wanted to discuss how this impacted fair use of the house, and would frequently digress into what the first group perceived as offensively irrelevant and time-wasting side issues. The levels of hostility that ensued were beyond anything I’ve ever experienced. They reverberated for years. I eventually started applying some of your heuristics (they are generally applicable!) and found the hostility greatly diminished.
Separately:
>This alone would explain why people doing torturous tasks should optimize against deliberation, for binary or multiple choice questions rather than open questions, for legibility rather than optimality. However, it is more generally true that the closer you get to collapse, the more urgently a survival-oriented system needs to act, the less capacity it has available to connect multiple pieces of information into a detailed model of what to aim for.
This sounds an awful lot like the way the left brain (as opposed to the right) operates, under the hypothesis proposed (at great length!) by Ian McGilchrist in “The Matter With Things.” The left hemisphere, under this hypothesis, is concerned with breaking things down into manageable/manipulable parts and is less good at integrating, perceiving a whole, and coping with open-ended questions. I mention this here because of the obvious brain-hemisphere context of the story (in an open-ended way: not suggesting that the location of the tumour “explains” your experience or theory). Also because, since contracting COVID March 2020 and suffering from long COVID over the subsequent four years, I have found myself struggling with deliberation and frequently experiencing something that I now see feels a lot like RAM overload — above a certain threshold of stress/choice complexity, my nervous system seems to refuse the task and “crash.” I preciously (tentatively) attributed this to neurological damage/inflammation from the virus — my symptoms during the acute phase were primarily neurological, eg altered taste and smell — but now I have an additional and highly interesting interpretation available.
Great post! Lots of thoughts. I think this actually has relevance to the Sorry You Feel That Way discussion. I was anti-SYFTW in that discussion, thinking about it in light of this dichotomy, when I'm in survival mode I think feelings and stuff are relevant pieces of information and I resent the implication by my thrive-mode opponent that they should be discounted. Thank You For Your Highly Enotional and Valuable Contribution? Hmm. I don't know. More martianese. The point is I get so hostile in those discussions that the formula my opponent uses will probably only make a marginal difference. But forewarned is forearmed. I do try and thank people for replies on here and other commenters do too and that seems good.
I hate to be that guy who points at ancient Greece, but I can't get rid of the feeling that within your framework, cynicism as a philosophy could be understood as a way to merge what you call survival orientation with thriving orientation. That is, trying to reframe the absolute minimum that allows you to survive as a way of thriving (in close connection with nature ofc). Funnily, what writings we have of cynics describe their style of teaching and behavior as blunt, raw, and often utterly disrespectful of social norms - suggesting that somehow, space-efficient communication "prevailed" upon the integration of the two systems. This may have something to do with the fact that cynics were not solving "problems" (I'm being unfair here, what you described is a tragedy, not a problem), rather, they communicated this way as they felt compelled to present themselves and their teachings to others. I was just wondering, are there other old philosophical systems we can think of that attempt to merge these two systems you talk about, and if yes, is there anything that 1) we could learn from them with regards to potential new "heuristics", and 2) can your theory and formulation add anything to these philosophies (either as an extension, or maybe criticism)? Cheers!
Improving on Cynicism sounds like a cool project, and doable since you can go far if you just quantify everything. But I think you can do that better than me, since you're evidently more familiar with the literature.
I would recommend stealing liberally from Eliezer Yudkowsky's Sequences. Because the Bayes Theorem does apply, everywhere. The old cynics didn't have access to it but you do.
If you believe in the multiverse, then you should feel assured that if survival is possible, you will survive (somewhere), even if the rest of us here observe your death. Please don't be afraid.
The survive/thrive continuum is immediately applicable to lots of things in my life. When I worked as a welder, my long-winded explanations of how to optimize production only annoyed my boss, who was being harassed by customers who needed their shipments yesterday. As I'm working as a farmer, my attempts to make precision agriculture software work flawlessly fall on deaf ears as my uncle steers the harvester manually in order to get the crop in before winter strikes. Urgency creates the survival mentality, to the potential detriment of long-term profits. This is a useful new perspective on the world.
So let's see if I got this right: When physical survival is at stake, trained professionals follow protocols which save on information but at the expense of extra time--or, to put it another way, the necessity of saving someone makes everything take longer.
On the other side of the coin: When thriving is at stake (that is to say, routine happiness), then ordinary people (or is it still trained professionals, just in a different profession?) follow routines that saves on time but at the expense of an additional information burden--or, to put it another way, we live life fast but talk a lot.
Is that right? If so, that's very interesting, because that's almost precisely the opposite of what I would have predicted.
I think "time efficiency" here is used in a loose sense.
You can think of "time efficiency" as "doing things the best possible way" and "space efficiency" as "doing things the way that uses the least resources".
The doctors and nurses in an emergency situation aren't usually concerned with applying the best possible treatments from cutting-edge research and using the best possible procedures. They are concerned with applying procedures everyone involved understands and can reliably execute in a repeatable manner with minimal space for error to come into the process.
The resources being saved here are each health professional's mental bandwidth. They could definitely probably do things in better ways, but that would carry an unacceptable risk of the whole process breaking down at some point because the people carrying it out aren't sufficiently familiar and "drilled".
So it's more a tradeoff of best possible results in exchange for maximizing reliability?
Well, best possible theoretical results, yes.
I feel like a more natural explanation to me would be that you previously overestimated the value of added detail.
Consider the 80:20 rule, that 80% of effects are due to 20% of causes. Applied to communication, you'd get that 80% of your message could be gleaned from 20% of your words.
In many cases, where accuracy is very important, spending the time to get 100% of the message may be valuable. (People who like reading several thousand word blog posts will likely appreciate this).
On the other hand, in a crisis situation, it's probably less worthwhile to spend 400% longer to provide just 25% more accuracy.
This is less like the Algorithms idea, since both space efficient and time efficient algorithms are *accurate*. It instead suggests that accuracy is worth trading off for time.
(And in a hospital, this may be the case. If you can attempt to cure 5x as many people, but lose 1/5 of them due to mistakes, then you're still up by 3. Though I doubt anyone thinks of it this way, since I'd expect a heavy moral reaction to this level of constant trolley problem. Also the numbers are probably different.)
Strong defense of this guy's thesis.
At many points in your life, you will be faced with a power system that has a different goal than you do. At some base level, it will have optimized around either survival or thriving, and so will you have done. Now what?
His seven heuristics are excellent professional and marriage counseling advice. Problems between people are almost always problems of varying values and goals. Learning how to talk about it is hard. Learning that skill opens new doors.
Several of my promotions and job opportunities have arisen from challenging an authority in response to a question, or telling them what I really believe and then asking them about what they believe. Often we're revealing staggering difference, but the journey of being vulnerable and then talking about something hard without guard up is revelatory to many people.
I often tell people ranting about someone that "there might be a hole in their map" in the same way.
p.s.
I don't mind guest posts if it doesn't impact your productivity.
Hosting others makes it more likely I'll defect to their views, though.
And here I was expecting you to provide an analogy between how humans make these tradeoffs and how tunable algorithms balance between these extremes. Like
Enhancement Heterogeneous Earliest Finish Time algorithm for Task Scheduling
https://www.cri.ensmp.fr/classement/doc/A-681.pdf
Fully Polynomial Time Approximation Schemes (tunable ε )
https://en.wikipedia.org/wiki/Fully_polynomial-time_approximation_scheme
or simpler Scapegoat Trees (tunable α)
https://en.wikipedia.org/wiki/Scapegoat_tree
I found the essay massively irritating and also very valuable. Thank you for sharing it. However, I have a conceptual disagreement.
Space efficiency seems to map poorly to the notion that the information needs to be compact because it might get overwritten. A space-efficient computation usually uses its limited space very intensively, rewriting the information in the available memory many times.
A busy nurse definitely wants a compact summary of the system state but also doesn't want to keep updating it continuously. Communication complexity therefore seems a much closer match, and is also a closer match to the information theoretic formulation.
Seeing like *Seeing Like a State*: death is highly legible while not thriving is not. In the US, we should also consider legibility to the law as well. Changing behavior requires changing incentives, not changing collaboration, as the incentives themselves ensure equilibrium. The _State_ goes just as far and wrong in the other direction, too, in state assisted suicide for depression or the like. There, the want of thriving is valued over survival itself, and its totalizing adoption is just what we ought to expect.
Huh. When my late wife was struggling with her serious illness, I didn't seem to have this particular kind of problem communicating; when I infodumped it often seemed to help. The situation was different in a few ways, though.
1) I wasn't the patient. For whatever reason, I was able to advocate for my wife more effectively than she was able to advocate for herself, and my wife would often ask me to try to convince the doctor or nursing staff of something.
2) My wife's condition was rare, to the point where most of the doctors and nurses treating her hadn't seen it before. If there's no standard playbook, when the patient's husband is telling you "get her back on this drug, it's important" you're more likely to listen.
3) We were in a hospital in the United States and not Germany. I do not know how much of a difference this makes, because I've never been in Germany.
4) My wife's condition meant that she had a series of extremely long hospital stays, and the nurses would get to know her pretty well over the course of a month or so.
5) Maybe I did have this kind of problem and didn't realize it. There were things my wife and I were concerned about that I couldn't get the hospital staff to act on; the doctor kept telling us not to worry, but at least one of them grew into a bigger problem.
This seems a useful framework to analyze many disagreements in the world (in politics, work, family, etc). Not sure how much the "thankful" recipe would help.
How this can viscerally feel from the inside as a thrive person: people are selecting from a menu of pre approved options. If your option isn't on the menu they will respond angrily that you must select an option from the menu. Importantly, the question 'what are the menu options' is not an item on many menus! It is easy to assume hostility because this process sometimes is actually hijacked by hostile patterns!
not only are liability focused processes playing not to lose, they treat more information as potentially more liability.
This is mostly orthogonal to the post other than the link to alphabetic sorting and computer science. When I was a grad student working as a teaching assistant a frequent task was sorting exams alphabetically by last name. First, the environment: you have yourself, a stack of unsorted exams, and a small table space in front of you.
My algorithm: Establish an alphabetical spatial bin system on the table based on the first letter of the last name in a (rough) grid pattern. Iterating over the unsorted stack, place each exam in the appropriate bin as received.
After all unsorted exams are binned, proceed to sort each bin. Most bins will be small and can be easily hand sorted without any special consideration. If a bin is particularly large, the algorithm can be replied recursively to the next letter in the name. Sorted bins can then be stacked (merged).
Justification for algorithm: your work space (table) is limited, so an algorithm like merge sort is impractical as you can't lay out all the papers in front of you. Even if you could, the time to physically access the papers on the table is lengthy and would negate any efficiency gains . Human short term memory is also limited and fallible so keeping track of the changing bins becomes problematic. A complex algorithm like quick sort would suffer similar issues, particular with regards to human fallibility and forgetfulness. The grid bin system works well with human memory: as you process exams, your muscle memory gets better at remembering precisely where each bin is, giving increasing efficiency gains to binning exams as you iterate over the stack, it also only requires processing a single letter at a time so mental processing is kept low, and the simplicity of the task allows all mental energy to be focused on that resulting in a low error rate. Last name letter distributions are somewhat uneven, this results in several large bins and many smaller bins. The smaller bins can be easily hand sorted by normal human sorting without any special consideration, there are often runs of these which can be stack to free space for recursively sorting the larger stacks by the 2nd letter. I don't recall ever having to go past the 2nd letter with the grid system for the typical input problem size.
This algorithm is called radix sort. It’s useful on computers too.
This does seem quite relevant to the medical field, comparing my own childhood and late teen experiences
I would not categorize this difficulty in communicating as a mismatch between systems attempting to “thrive” and systems attempting to “survive.” I’m a lawyer and I encounter the same communication problem all the time.
The mismatch the author encountered in the hospital setting is between communicators (like the general public) who are not members of a particular *profession* and persons who *are* members of that profession (in his case, between himself and healthcare professionals).
For example, people come to me with what they believe might be a legal dispute. My first job is to determine: is this even a dispute? If so, is it a dispute that lawyers can solve?
In the course of my conversation with the potential client, the potential client will offer a *lot* of information. People generally do not know what is relevant to a lawyer, so they just pour everything onto the table. Quite often, initial meetings with a potential legal client resemble a psychiatric session more than a meeting between lawyer and client.
But, as a lawyer, I’m trying to extract a small kernel of “legally relevant” information from this outpouring of (mostly irrelevant) information. And I get impatient. I try to focus the person on the information I need. But of course, it’s hard for my non-lawyer client to stay focused on “legally relevant” information -- they don’t know the law. They don’t *know* what’s legally relevant. So, the potential client is constantly veering off course, into irrelevant stuff, and I’m constantly (somewhat impatiently) trying to steer them back to what’s relevant.
Now, sit me down with an engineer who is trying to extract from *me* “information relevant” to a possible engineering problem. The engineer has the same communication problem with me that I, as a lawyer, have with potential legal clients -- I don’t necessarily know what is relevant to an engineer. So, I’m just dumping information on the (increasingly impatient) engineer, who is trying to extract a small kernel of “engineeringly relevant” information from me.
This seems to be the same problem the author encountered in the hospital. In the hospital, healthcare *professionals* are speaking their own professional language, just like lawyers and engineers do within their respective domains. The healthcare professionals are trying to extract “medically relevant” information from the author -- but the author doesn’t necessarily know what is “medically relevant,” so he’s constantly veering off course as (increasingly impatient) healthcare professionals try to get him to zero in on the medically relevant information they need in order to discharge their professional obligations.
Again -- the mismatch in communication systems is not a mismatch of “survival systems” versus “thriving systems.” It’s a mismatch of “non-professionals” trying to communicate with professionals *within* the professional setting. More simply, the communication problems arise because the non-professional has a weak understanding of what information is RELEVANT to the professional, in order for the professional to carry out his or her professional responsibilities.
More generally, I’d say the author has discovered that different groups (I’m now generalizing beyond the word “professionals”) have different understandings of what information is relevant to *them.*
It’s probably not relevant to an astronomer that you ate Cheerios just before you spotted a new, previously unknown comet. But hey, maybe, as it turns out, that *is* relevant information! Maybe astronomers (and the rest of us) have yet to learn the connection between eating Cheerios and spotting new, undiscovered comets!
The point being that, once you get into exploring communication problems based on relevance, you realize that what is relevant to an in-group will shift over time, as new understandings emerge.
The author did say that the professionals were overlooking information that was relevant to them that they presented, which caused avoidable problems.
Relevancy for different systems is important when exchanging information, but I see that as a subset that is addressed in the author’s heuristics.
The post’s thrust isnt that information exchange errors exist and that these are the strict types of errors when exchanging information, it’s about offering principles which overcome generalized communication errors between systems. So I dont know if finding examples of communication errors that don’t fit neatly into the types disqualify the heuristics’ validity.
As a nurse who went from signal/survival systems to noise/thrive systems (in other words, primary care to psychiatry) this tracks. It's why primary care doesn't GAF about emotions, or the unconscious, the system does not have time. This is fragility/anti fragility in other words.
I enjoyed this and found it useful (I’ve always struggled to communicate with most people). Thanks for posting!
I can believe that there's an insight here that is of great importance, but I don't quite see it in this post! (Understandable—the essence of an insight into a deep fractal tension can be hard to convey, speaking as someone who has had a couple.)
> There are existing computer science results about how to integrate space-efficient and time-efficient algorithms. Whatever the specific practical details of the malfunctioning collaboration that you want to improve, these results will be applicable. This is the principled solution!
I could look up the CS results I guess, but I'd be curious for your take Daniel on how exactly the math there bridges to the 7 heuristics, which mostly to me feel like pretty generic takes on how to navigate two different ways of doing things, not a profound insight into how to navigate the dance of these particular fundamental types.
I'd like to see:
1. some specific stories of how you applied these insights, in both roles
2. a brief explanation of the CS results and how the heuristics map onto them
Love the theory and approach. The fact that it's using computer science and applies it to human communication might make the latter more efficient and error proof. I wonder if it can be applied to political discourse, like for Presidential debates or to argue with a person from an opposing party. What I fail to grasp is who is who on the continuum of survival-thriving, i.e. are Democrats more concerned with survival and Republicans with thriving? Maybe not, or not for all demographics. I would say that Democrats are more focused on the low-middle class, while Republicans are protecting the Upper/Ultra rich. Maybe from this perspective we can use the theory proposed by the OP - the Rich will suffer if the poor will get too poor (or the middle class would substantially shrink), and the poor/middle class will not be able to thrive as much if there is no wealth generated by Industry and Innovation (cash hoarding and pure greed are not part of Innovation cycle) which would fill the coffers and fund the social programs. This dichotomy has been somewhat resolved/improved upon in Europe, especially in Scandinavian countries, but USA seems to be very stubborn to negotiate on the mutual goals. The Taxation and Social Programs expenses are both very heated subjects. One side wants to Tax the rich to the max, while the other side wants (and is somewhat successful) in not paying their share of Taxes. We need a win-win solution that would satisfy both sides.