Highlights From The Comments On Elon Musk
[original post: Book Review: Elon Musk]
1: Comments From People With Personal Experience
2: ...Debating Musk's Intelligence
3: ...Debating Musk's Mental Health
4: ...About Tesla
5: ...About The Boring Company
6: ...About X/Twitter
7: ...About Musk's Mars Plan
8: ...Comparing Musk To Other Famous Figures
9: Other Comments
1: Comments From People With Personal Experience
As somebody who also worked on getting humans to Mars (the Orion project, which now only going to the moon after a large demotion), yeah having good ideas at those companies is soul crushing.
Getting trivial changes done to anything takes 6-12 months. I’m talking 1 hour fixes. Because they weren’t planned for already, so they can’t even be planned in this 3 month cycle.
And on the workload front I used to put in 70+ hour weeks every once in a while, so working for SpaceX would be a large step up in quality of getting things done and getting to build the cool stuff, while not being a horrendous downgrade in the other dimensions (though 75 hour weeks eat you alive. It’s basically 6 hours of sleep per night and every other moment is reserved to working or getting ready for work / commuting).
Fluffy Buffalo answers:
I understand the frustration... but my impression is that space exploration is one of the fields where very thorough, very systematic planning with very conservative change cycles is the most promising approach to get something that works at the first attempt - even if it takes longer and costs more than planned. Compare the JWST to the most recent "Starship" launch for illustration.
This would sound plausible, except that Musk has succeeded by doing the opposite. I think this is why so many people are in love with Musk: he’s proven that valuing good ideas, moving fast, and not having bureaucracy can work, sort of, in a weird little bubble of his own creation. Yeah, the first Starship exploded, but most people predict Starship will eventually work, and when it does it will be a much more impressive feat of engineering than JWST or anything else created the “proper” way.
I’m not sure whether this means that everyone else is an idiot with a pointless bureaucracy fetish, or that only a few very special people like Musk can make the non-bureaucratic version work.
I have worked in space engineering for years and my impression is that the big space agencies and companies have a lot of inertia and reluctance to consider new ideas and change. A lot of the processes have been built up in response to past failures, but they also stifle a lot of innovation. When people come in with a fresh approach and the resources to implement them, they have tended to get quite far. You can look at how SpaceX has done, but also the early days of the space program at NASA were a lot more open to innovation than today.
He is a polymathic engineer of rare but not unparalleled breadth, and the sequence of areas he attained expert-level knowledge in, and hands-on skill, are a first-order factor in the probability of success of the series of companies.
— Software first.
— Off the shelf hardware in parallel (the datacenters) with scaling
— Hand-built maximally esoteric hardware second (Falcon 1)
— Then hand-built medium-complexity hardware with a firm eye on eventually manufacturing at scale (Roadster).
— Then medium-complexity manufacturing at scale (Model S).
1. Building expert level skill creating software takes longer, and is less likely to occur than any other skill required in any of the industries that Elon‘s companies operate in. It is basically impossible to become a world-class software developer if you start after you’ve achieved career success in another industry.
- 1 is hardest, and most important
2. Hands-on expertise in spaceflight physics, metallurgy and fabrication, rad-hardening, rocket engine design, spacecraft structures, NDE/NDT fixturing: the physics and builder-level skills for these can be learned on the job and with intense solo study by a sufficiently motivated and adequately intelligent person within a few years, if given a free hand to roam/rotate.
- 2 & 3 are harder than all that follow
3. Space tech is nearly maximally esoteric, engineering and construction-wise. (Only the largest multi billion $ physics/astrophysics projects have a larger design envelope than space tech). Building expertise in space tech makes terrestrial engineering and fabrication challenges like car parts and solar panels seem pedestrian by comparison.
4. The techniques for manufacturing macro scale components (everything larger than 1mm) at scale, and subcomponent assembly at scale, etc etc - these can also be learned on the job a sufficiently motivated and adequately intelligent person within a few years, and some mfg folks to absorb knowledge from.
5. Learning supply chain optimization is something 1/4 of humans can lead to do well, 1/20 can learn to do well in their spare time, and is slightly more than a triviality for anybody who can handle items 1-4.
I’m making these positions from a position of having some experience in all five of these areas, for satellites, rockets, cars and other vehicles, but starting with software. Biased but also have trod the path in the same sequence, just not nearly at the same level of success, and I firmly believe that the sequence matters.
See more discussion here.
I remember the days when SpaceX was really ramping up university recruitment. They were the table at the career fair everyone wanted to give their resume to. Naturally, SpaceX sent a spectacular a-hole who yelled at and belittled most of the students applying. It got so bad they actually apologized about it when they held a talk at the next career fair. Turned quite a few folks off, it was a real embarrassment. Took a couple of years to wash that one out.
I sometimes think about the people that knew the type of behavior going on and still stood in line and applied. I think a large part of why all those ex Musk employees and etc. still excuse various behaviors and defend him so fervently is that there is approximately no one who goes to work at one of his companies just to work a job. No one would put up with that crap for a 9-5, and now that it's so well known, no one would apply for it. It's all starry-eyed (mostly recent college grad) true believers. And the turnover rates speak pretty well for themselves […]
Some more thoughts after sleeping on this review. It's very strange... so being an automotive engineer for several of those "staid, evil" Big3 companies, one gets a very direct view of Tesla and how they have been over the years.
Something that probably ought to get talked about more: for large companies, we are among the first few hundred to buy the newest hotness from our competitors. I saw a Model X Founders' Edition fully disassembled on tables, with the welds drilled out and sectioned so we could see every single part. I've done side-by-sides with Teslas and various other vehicles, where we literally will put our part and the competitor part next to each other in a giant warehouse (all of them for a series of vehicles) and do side-by-sides. When you do that, abstract questions of genius kind of fade to the background, and you get to actual real world questions like "is this part good? Is it better than mine? What is it trying to do? How does it try to do them? What does this say about the engineer's constraints? What does this say about the company organization behind it? Where are the organizational seams? Where are the hard points that could not be changed? How do those reflect on my company, my program, what we're trying to do and the things we have to work around?"
This isn't just idle navel-gazing. Akins' Law about system interfaces is quite relevant here. Where you draw organizational and system boundaries and the restrictions you put on certain hard points can drive significant differences in a component on a table.
But out of all of that, my biggest take-away was that Teslas..... just aren't very good? Their structures up to the Model 3 are quite inefficient and don't have great rigidity. The dimensional variation is shocking (far beyond even SBU, IYKYK). The hang-on parts are generally relatively poorly performing on their own. They can't touch our structural or powertrain durability tests. Rate and handling is bad, ergonomics fails to meets package targets, NVH and sound quality are poor, and we pay JD Power far too much to find out just how bad the quality numbers are (hilariously bad). I don't think it's an exaggeration to say that most other OEMs can't make a Tesla, because our systems and processes prevent us from releasing something that half-baked.
It really makes you question the customer sometimes, because if we put out a touchscreen that failed like that, we'd rightly be ridiculed. CEOs have lost their jobs over far less.
I think Musk's genius is in two very closely related areas: getting investors to give him an unlimited checkbook, and in getting customers to believe they're doing something new, novel, and important, in a way that lets him walk past screwing up things that legacy players get right as an inevitability. The technical side? Most engineers I've met can probably accomplish it.
P.S. the interface is so slow and laggy, holy cow
Paul T writes:
>> “Since these companies already have hundreds of engineers, each specializing in whatever component they’re making, why does it matter whether or not the boss is also a good engineer?
Part of the answer must come from that story above about him taking over people’s jobs. His strategy is to demand people do seemingly impossible things, then fire them if they fail. To pull that off, you need to really understand the exact limits of impossibility.”
I agree with this, and would add that it's not specific to his strategy of micromanaging folks and taking over their audaciously-scoped tasks if they can't complete them. For any tech company, a technical CEO has superpowers compared to a non-technical one (but are predisposed to a fairly standard set of weaknesses too).
Even beyond just the CEO role, in the tech industry there is a very widely discussed challenge of "technical vs. non-technical managers". The engineers doing Individual Contributor (IC) work can grow to resent non-technical managers if they don't have a sense for how hard a given ask will be to implement, and a common anti-pattern is for the non-technical product, marketing, sales, and scheduling decisions to be made without a deep understanding of the actual feasibility as it bottoms out in the technical implementation. At worst this can lead to myopic leadership ("MBA management" etc.).
At the end of the day, a non-technical manager/CEO must be good at synthesizing the team's estimates and opinions, and knowing when to defer to concerns about tactical considerations, vs. take a tactically more difficult path which will advance strategic aims. (Aluminum chassis is a great example of this kind of tactically-painful but strategically visionary decision where a non-technical leader might struggle.) In a normal tech org the CEO has to trust the CTO, and then the CTO works through layers of managers to enact their technical vision, so there are multiple hops where the CEO's vision can be lost in translation. At Tesla Musk is collapsing both CEO-CTO and CTO-manager-IC communication down to him directly talking to ICs, which (while having other obvious organizational issues) allows him to make bold technical bets and stay very aligned on what is actually possible for his ICs to do.
Coming at the same issue from the bottom-up direction, technical ICs often don't have the context of the full strategic vision, and non-technical leaders often struggle to communicate it downwards in ways that are meaningful to the technical implementors. This is another thing Musk is better than almost anyone at; taking a lofty objective and chaining it down to an individual's role. I heard a SpaceX employee giving an answer in an interview like "Our mission is to become an inter-planetary species. To do that we must first colonize Mars. To do that we need to build a heavy lift rocket (Starship). To do that we need to build a more powerful engine. To build our new engine we need this valve assembly to work; my mission is to optimize this valve to X performance requirement".
Having said all that, why not just use technical managers? The answer is that it's usually not the best use of a strong IC's time; managing is very hard, requires strong empathy, is hard to teach, and training is criminally underfunded and under-appreciated. Managing is very different than IC work; it's meetings and interrupt-driven communications and performance management, whereas ICs usually thrive on "Maker Time" where they (optimally) get long blocks of uninterrupted time to get into the flow state and think about one problem. So while a good senior IC starts to get involved in communications and scheduling and other "outwards-facing" non-technical activities, there isn't an obvious universal progression from IC to manager. It used to be quite standard to have "senior IC" as the pinnacle of technical career progression, and the only way to get promoted further was to become a manager; this turns your best ICs (technical leads, mentors, or whole-system generalists) into normally-distributed managers (i.e. some good some bad, with no expectation for them to be better-than-average). Now at least in software it's more common to have a strong IC progression track that's parallel to managers, but you still see some degree of "strong IC -> mediocre manager" career paths.
The reasons you'd favor non-technical managers also apply to why non-technical CEOs are usually better at their jobs; in most organizations, the technical work is one or maybe a handful of roles in the C-suite (you might have a CTO and a Chief Scientist, say), while there are more non-technical roles (Sales, Operations, Marketing, Legal, HR, fundraising, and so on), and the CEO needs to be something of a jack-of-all-trades between all of those; in aggregate, non-technical skills are required more than technical ones. Musk's successful companies are outliers in that they benefit from being heavily technology-focused; they are applying tech company style iterative innovation and experimentation to historically non-software/non-"tech" domains, which I believe increases the importance of the CEO->CTO->IC chain, and is why Musk's strength in that area is disproportionately impactful. Having a Musk-style technical CEO would not be useful in a traditional car company, or a sales-driven enterprise software company like SAP.
2: Comments Debating Musk’s Intelligence
Can you think of any other reasons why graduates of elite universities might describe their famously vain and petty boss as very intelligent on the record other than assuming he must be a 1 in a million intellect?
Liminal Revolutions (blog) writes:
I have seen several interviews with Musk and in none of them does he come off as being >120 IQ. Second-hand reports claim that he is very smart but I don't believe second-hand reports. Does anyone have any unfalsifiable proof that Musk is highly intelligent, like a video of him giving an unscripted talk about numerical methods for control systems or something? If he is able to meaningfully interrogate his employees about their work he should be able to lecture about various aspects of aerospace engineering. Where are the videos of him clearly demonstrating a genius-level IQ?
Schroden Katze writes:
I seriously dropped my estimation after the Python script saga when allegedly AI-interested guy and former programmer didn't even comprehend what is Python script
I blow a whole foghorn in attention when a guy who looked smart for me fails precisely at point I know about
I'm pretty stupid, but I wonder if the people that are being interviewed and speak highly about Musk are trustworthy when they talk about him.
I mean, Vance says this: 'When Musk learned he was being profiled, he called Vance, threatened that he could “make [his] life very difficult”, ...'
Why wouldn't do Musk the same to people who dare to threaten the narrative of him being some high IQ physics & engineering genius? I haven't seen any direct evidence of that to be honest, it's rather the opposite actually.
Hmm. I was glancing at the new Isaacson biography. One of the sources is Max Levchin, another PayPal executive. Max is now a billionaire himself, and seems mostly retired. As far as I know he doesn’t depend on Elon for anything, and he’s rich enough that he would be hard to threaten. And he was mostly on Thiel’s side against Elon, and tells a lot of stories of Elon making stupid decisions at Paypal (which he explicitly calls out as stupid decisions). So he’s about as fair a source as we’re going to get. Still, he has stories like:
And yet, Levchin began to marvel at the counterexamples [to his generally low opinion of Musk], such as when Musk astounded him by knowing things. At one point, Levchin and his engineers were wrestling with a difficult problem involving the Oracle database they were using. Musk poked his head in the room, and even though his expertise was with Windows and not Oracle, immediately figured out the context of the conversation, gave a precise and technical answer, and walked out without waiting for confirmation. Levchin and his team went back to their Oracle manuals and looked up what Musk had described. "One by one, we all said, 'S**t, he's right," Levchin recalled. "Elon will say crazy stuff, but every once in a while he'll surprise you by knowing way more than you do about your own specialty. I think a huge part of the way he motivates people are these displays of sharpness, which people just don't expect from him, because they mistake him for a bullsh***er or goofball."
I’ve heard via my personal network (which I trust more than this Ashlee Vance book) that Musk did used to drill down into engineering level decisions at SpaceX. Whether he’s actually extremely technically proficient, I don’t know. I’ve also heard this has slowed down a lot in the last few years since he’s focused on Twitter. (Note that the above is hearsay, I’ve never met the man myself or worked for any of his companies).
I don't think most people know the general state of aerospace industry CEO's and managers. Boeing is currently run by a former hedge fund guy with a degree in accounting. Relative to the industry, Elon Musk's public statements demonstrate enormous engineering acumen for an exec. I've got a master's degree in AE and whenever he's said something in aerodynamics or structures I'm like yep, that's about right. I remember a while back Musk said something about the 787's batteries catching on fire and some MIT prof, world expert on batteries, was quoted saying, "I would have said exactly the same thing."
I've worked along former SpaceXers and hung out with current ones (mostly in outdoors sports). If you work in the industry, especially in LA, you run into them. I was also interviewed by Brogan at Hyperloop a while back (super nice guy). The SpaceX hiring bar for technical talent is super high and I wouldn't exaggerate to say the average SpaceX engineer is twice as talented and hardworking as the average Boeing guy. Also, pretty arrogant in my experience (versus Googlers I've met tend to be humble even if they went to Stanford). I think this really started from the top of the company and he couldn't have built this pyramid of insane talent if he didn't have an informed, critical understanding of mechanical engineering.
heliotropic on Twitter writes:
someone close to me worked w elon and said the same - he’s breathtaking at instantly understanding technical problems & coming up with solutions a room full of phds hadn’t considered.
The new Elon Musk biography says that Musk’s SAT scores were 730 math, 670 verbal. This are good-but-not-great scores now, but epursimuove reminds us that Musk took the SATs before 1995, when the scores were calculated differently and high scores were harder to get. Based on the pre-1995 norms, 1400 puts him at the 99.1st percentile of test-takers. But only a third of students took the SAT during 1994 (the year I have good data for), and probably these were selected for intelligence.
So Elon is somewhat higher than the 99.1st percentile of the general population. I think this suggests a total IQ somewhere between 136 and 140, probably around 130 verbal and 145 - 150 performance/math. Sorry, making the adjustments described here, it’s more like 130 - 135 total, 135 - 145 math.
You can always argue a bigger and bigger conspiracy. Maybe Elon forces employees to tell their friends that they think he’s smart. Maybe he has some kind of blackmail material on Levchin. Maybe he suppresses publication of books that contain anecdotes proving him to be stupid. Maybe he lied about his SAT scores to Isaacson (I don’t know if Isaacson has any corroborating evidence).
But I think our prior against him having very high (though not best-in-world or unprecedented) engineering ability just shouldn’t be that high. Suppose I told you about “my friend”, “Eli”:
His father was a successful engineer who became a multimillionaire from his engineering business.
He was obsessed with engineering as a child and studied for hours with his father every day.
He got accepted to a Stanford PhD program in engineering
He dropped out to start an Internet startup, exited successfully, and now he works at SpaceX.
Not knowing anything else about this “Eli” guy, if I told you he was also a really amazing engineer, you would say “yeah, of course”, and accept it as a pretty plausible thing for someone with this background to be. Surely adding the information that he made $200 billion and ran Tesla along the way shouldn’t decrease our credence in this possibility!
adderallposting (relevant name?) accepts my estimate of Musk’s intelligence but challenges my estimate of his intensity:
I really, really, doubt that there are only 30 people in America more intense than Elon Musk. The biography, which was clearly trying to paint a picture of Elon as a particularly intense person, made him out to be more of a 1-in-10,000 intensity-level person than a 1-in-10,000,000.
If Elon was 1-in-1,000 intelligence, and was merely 1-in-10,000 in terms of intensity, and intensity was completely uncorrelated with intelligence (I would tend to think so, roughly speaking - plenty of dumb people have hyperfixations, too) then of his top-300,000-most-intelligent-Americans cohort, he would be in the top 30 most intense, or alternatively, for his top-30,000-most-intense-Americans cohort, he would be in the top 30 most intelligent. This, combined with the luck that Scott mentioned in the few paragraphs preceding the text from the OP I quoted, seems a combination completely sufficient to explain Elon Musk's extreme success even in terms that are still almost maximally generous to the 'Elon is successful largely because of his particularly effective personality' theory.
This is a good point! Probably there are very intense schoolteachers, plumbers, and bank tellers, but we never think about them!
3: Comments Debating Musk’s Mental Health
Surprising to see a psychiatrist write a review of Musk focusing on his psychology and replete with quotations about his erratic sleep habits or obsessive focus, and never use the words "bipolar" or "mood disorder"
The link goes to many articles that Gwern thinks provide evidence, including some where Musk self-describes as bipolar (“maybe not medically tho”).
Musk’s ordinary behavior - intense, risk-seeking, hard-working, grandiose, emotional - does resemble symptoms of hypomania (full mania would usually involve psychosis, and even at his weirdest Musk doesn’t meet the clinical definition for this).
But hypomania is usually temporary and rare. A typical person with bipolar disorder might have hypomania for a week or two, once every few years. Musk is always like this. Bipolar disorder usually starts in one’s teens. But Musk was like this even as a child.
Musk does describe sometimes having “terrible lows” and taking ketamine for “depression”. He doesn’t say if this one was diagnosed, but I’m a little skeptical. While granting that he has extremely bad times, an official depression diagnosis requires symptoms other than low mood, including things like fatigue, feelings of worthlessness, loss of interest in doing anything, moving slowly, cognitive impairment, and even suicidal thoughts. More important, it requires that the person not have hypomanic symptoms at the time. Also, it has to last at least two weeks, pretty constantly. Do we have evidence that Musk has been fatigued and felt worthless and just wanted to lie around in bed and not cared about Mars or anything for two straight weeks? I don’t know, maybe he has! I just don’t think the fact that he’s “haunted by demons” and sometimes goes into rages necessarily establishes this. I would guess that Musk’s bad moods happen when something is going wrong in his life, last days-to-weeks instead of weeks-to-months, and are marked by high energy (even if that’s angry/nervous energy). These could be any of a number of things - borderline, autism, narcissism, normal bad moods - but they don’t really match depression.
His low periods might meet criteria for a mixed episode. But a bipolar disorder that starts in childhood, continues all the time, has no frank mania, and has only mixed episodes instead of depression - doesn’t really seem like bipolar disorder to me. I’m not claiming there’s nothing weird about him, or that he doesn’t have extreme mood swings. I’m just saying it is not exactly the kind of weirdness and mood swings I usually associate with bipolar. I have never met or talked to him and he probably keeps a lot of his inner life secret so I could be wrong, I’m just not seeing obvious evidence for this.
Musk’s intensity and energy sound more like hyperthymic temperament, a technical term for “kind of like hypomania but that’s just how you are”. Wikipedia describes the following features:
increased energy and productivity
short sleep patterns
vividness, activity, extroversion
tendency to repeat oneself
breaking social norms
very strong libido
love of attention
low threshold for boredom
generosity and tendency to overspend
cheerfulness and joviality
irrepressibility, irresistible, and infectious quality
I think I would describe this as “Bipolar proves that a certain set of traits form a package that can come together, Musk does have this package of traits, but he doesn’t have it because he’s bipolar”.
Gwern is still not convinced:
Did you hear about Kanye West's depressive phases? I take it you did not, or you would have responded 'yes, I totally knew Kanye was bipolar before he announced his diagnosis, thereby showing I can in fact detect celebrity bipolarity before they are publicly revealed and thus it is meaningful that I don't detect it for Musk'. How would you know?
Unlike the manic phase where Kanye is ranting about world jewry and overruling his PR handlers, during the depressive phase, they do... nothing. That's kinda the point. They have PR handlers and buffers of stuff in the pipeline and the ability to make news by tweeting something dumb which takes 5 seconds and even a depressive can manage that much effort. From the outside, celebrities go dark all the time. They're on vacation, or they're heads down working on a secret project, or they're buffering news, or there's just Poisson clumping in what's going on, or news about them happens to not go viral. Or... they're depressed. Regardless of the cause, 'out of sight, out of mind.' You don't notice what you don't notice.
How would the world look any different to you than it does now if Musk did have depressive phases lasting a month, of the sort he self-medicates with the ketamine he's known to use (not to mention whatever stims or drugs he may be self-medicating with), where he mostly shit-tweeted while folks like Gwynne Shotwell continued to run SpaceX and Zach Kirkhorn ran Tesla (as they always have while doing their best to stop the techno-emperor man-child from follies like rolling out the next Tesla car without a steering wheel because 'FSD is going to work real soon now')?
[…] [We have strong evidence for Musk’s bipolarity], which I have put together in one place, from his family background to the unique demographics of bipolar entrepreneurs to his delusional beliefs & signaturely bipolar actions like trying to broker peace with world leaders to his hero complex to his own tweets about bipolarity to confidants hinting at the depressive phases to his known drugs of choice, and so on and so forth.
Psychiatrist here. [W]ithout psychosis, no matter how subtle, or some loosening of thought process, those symptoms in isolation don’t really raise flags for bipolar. His obsessiveness as you pointed out is focused and goal-directed. This would not be consistent with a manic episode. I’ve seen nothing from him that makes me think he is bipolar […]
Even hypomania shows some loosening of thought process and idiosyncratic changes to thought content. I think hyperthymic is the word OP is looking for. Not really considered an illness per se. Also I haven’t seen evidence for clinical bipolar depression as well. I obviously haven’t met or observed Mr Musk but any version of bipolar would be pretty low on my differential.
I find when laypeople say bipolar they often mean “moody” which is more consistent with a personality disorder.
[…and] when I mean idiosyncratic I mean idiosyncratic for Elon. So if he presented as very excitable, with a deceased need for sleep for many days without drugs/caffeine and started talking about how we all needed to avoid technology and social media and focus on god, then I’d be more concerned.
I notice the non-psychiatrists (including very smart people I usually trust) lining up on one side, and the psychiatrists on the other. I think this is because Musk fits a lot of the explicit verbally described symptoms of the condition, but doesn’t resemble real bipolar patients. You can decide how you want to classify this.
4: Comments About Tesla
Michael Watts makes an economic argument that Musk is overrated:
If we look at current market capitalization, Ford (stock ticker: F) is worth $50 billion. That is the price of owning all of Ford if you could buy all of the stock at current prices, which you can't do.
Meanwhile, Tesla (ticker: TSLA; this makes me wonder if Elon Musk is trying to get T) is worth $840 billion by the same metric, or just under 17 times as much. I am much less confident in the sales figures I pulled off the internet just now than I am in the market capitalization numbers, but they tell me that last year Ford "sold" 4.2 million cars and Tesla "delivered" 1.3 million, or 0.31 times as many cars as Ford. (I'd really like to compare number of cars manufactured, but good luck figuring that out.)
If neither company had outstanding stock and therefore they both had a market capitalization of zero, how likely would you be to conclude that owning Ford would be 17 times worse than owning Tesla?
If you buy Ford you also have to pay off its debt which makes the ratio a little less crazy. "Ford Motor long term debt for the quarter ending June 30, 2023 was $93.895B, a 10.45% increase year-over-year." vs. "Tesla long term debt for the quarter ending June 30, 2023 was $0.872B, a 69.91% decline year-over-year."
Using debt + market cap we get a ratio of ~6 instead of 17 (which is of course still substantial, but quite as crazy). Also the trends in sold/delivered debt/profitability definitely favor Tesla over Ford which is prob. worth something (though whether it's 6x or not is another question).
CY Hollander writes:
[While] Ford is a mature company founded 120 years ago, Tesla was founded 20 years ago and has been growing [rapidly] ever since. The difference is very material to any valuation of the two companies . . . estimating Tesla's "true" value at [Ford's market capitalization] * [Tesla's sales volume over the past year]/[Ford's sales volume over the past year] would have demonstrably and dramatically undervalued it for pretty much every moment in Tesla's history, by not accounting for the difference in growth rates.
5: Comments About The Boring Company
The Boring Company got brought up as an example that Musk doesn’t bat 1000. But is it a real failure? It’s valued at $6 billion and working on digging a tunnel underneath Las Vegas. Commenters weigh in:
AFAICT, the Boring Company has made tunnels cheaper by making them smaller and skipping all the parts that make them safe to use.
» "... and by skipping all the parts that make them safe [for internal combustion vehicles] to use."
Not an expert here, but my understanding is that you have two options:
1. Make a big tunnel to run a train through. This will cost billions.
2. Make a small tunnel to run cars through. This still costs a lot because cars have exhaust that needs to be cleared.
By running electric cars through his tunnels, Musk routes around both of these problems.
Paul T writes:
[Elon] was initially bullish that they would somehow be able to get a 10x cost improvement in tunnel boring machines, claiming to have found room for 4x improvement in speed from a thermodynamic first principles analysis, but he does not seem to have achieved much. (I was fairly skeptical that there would be that much free lunch, but he did manage a 10x cost improvement with SpaceX so I felt a strong inclination to defer to him at the time.)
Regarding the valuation, as the Twitter acquisition showed, he can text his buddies and raise billions. I don't index too much on Boring's $6b valuation; it just means he persuaded VCs to pay $675m for ~11% ownership stake during the zero-interest rate period, it doesn't say much about the company's ultimate profitability. For non-Musk founders I'd say that's a strong signal, but I think we've established that there's billions of dollars available for whatever project he's working on, with sparse due diligence, based on his reputation earned with Tesla and SpaceX.
I'm a tunnelling engineer in California, very pro-public transit, very anti-Musk for about 8 years now.
The California High Speed Rail project will be the fastest train speed anywhere in the US. In my view it's an exciting new type of project we haven't built before. But in public opinion it's panned as a disaster or "boondoggle" project. Yes I admit it has schedule problems and cost overruns, and this is a legitimate gripe about the project. But this is unfortunately normal for a construction project in California, or the US in general.
The Boring Company Las Vegas system is tunnelling a ~14 ft diameter tunnel that can fit 1 lane of car traffic, and it's light on some safety features like ventilation, exit walkways, or fire suppression systems. It will use Tesla cars, driven by Tesla employees. In my view this is basically an underground Uber system, but it will probably have more expensive fares to regain the capital costs of building the tunnel (Boring Company is paying for the tunnels, and casinos are paying for the stations, they do NOT have funding from City of Las Vegas AFAIK). But this expensive Uber system is exciting??
I think some observers see the situation as "The City of Las Vegas agrees that this company is legitimate" when really it's more like "okay, sure, we give you permission to build us a bizarre gadgetbahn system on your own dime, good luck".
Maybe California HSR fails, maybe Boring Company fails, maybe California HSR succeeds, maybe Boring Company succeeds. But I feel that the public is putting points on the scoreboard before the projects are completed.
Also, on a more abstract level, why would anyone trust a private company to make good public transit? If your transit system can make 20-30% of its revenue from fares, that's considered a win. Most of the budget is funded with tax dollars because transit is considered a public good. I agree that private companies can be more efficient than government, but public transit seems like an especially bad industry for a private company - they would need to charge sky high fares to regain capital costs, and they don't have eminent domain powers either. (Eminent domain probably isn't needed in Las Vegas - the casino owners, with large properties all along a single line, will probably be on board. But the plan doesn't scale to other cities.)
Paul T writes:
[Musk] did also posit big cost savings beyond the smaller tunnel diameter, and I think this might be a synergistic play; if we can build lots of small cheap TBMs then it becomes more viable to build lots of tunnels, not just because smaller tunnels are cheaper. Right now there are only a handful of the largest-size TBMs in the world, and I suspect if you pay all the freight cost to transport a TBM to your project, and pay the scheduling cost to wait for the next availability slot, you may as well get a big tunnel out of it. Vs. if TBMs are as common as backhoes, it becomes viable to do a bunch of stuff that simply wasn't cost-effective before.
(This is, broadly speaking, the SpaceX and to some extent Tesla playbook too, FWIW. Bringing economy of scale and turnkey solutions to a previously expensive bespoke project industry.)
6: Comments About X/Twitter
>> “He fired 80%-90% of the workforce [of Twitter] without any clear change in user experience. This was bad for the fired people and bad for PR. But it makes him look more competent than whoever was there before him and hired 5-10x more people than they needed.”
Two things to remember about this:
1. Twitter's revenue is mostly advertising.
2. Per Musk, advertising revenue is down 60% since the acquisition.
Musk says the lost revenue is primary due to pressure by the Anti-Defamation League (https://twitter.com/elonmusk/status/1698755938541330907), but I feel like "there are now a tenth as many people working there" is also a plausible hypothesis.
This is a great point. If you think of Twitter’s key job as attracting and keeping advertisers, then the user experience is just one part of that.
This doesn’t even have to contradict Musk’s ADL claim! If Parag or whoever employed a thousand censors to keep the ADL happy, and the ADL becoming unhappy cuts Twitter profits by 60%, then there’s a strong business case for those censors!
Jack Johnson points out that this is kind of a “paying the Danegeld” situation. But you don’t get to call someone a business genius for deciding not to pay Danegeld unless they show success in getting rid of the Dane.
FionnM argues that advertisers started boycotting the platform the moment Musk got it, based on Musk previously saying he supported free speech, so none of his actions as CEO can have caused the boycott. But even if this is true, maybe “him saying he supported free speech” can be interpreted as “him telegraphing he would do something like this”, and if he later proved that he wouldn’t, the advertisers would have come back.
There’s a much longer comment thread about these issues starting here.
My only gripe with the whole piece is this section:
>> “You could argue Musk’s first year at Twitter has actually had a lot of positives. He fired 80%-90% of the workforce without any clear change in user experience. This was bad for the fired people and bad for PR. But it makes him look more competent than whoever was there before him and hired 5-10x more people than they needed.”
Twitter has had more problems in the last year than I've ever seen it have and the moderation has been noticeably worse. There are reports of unpaid salaries and rents popping up here and there - no proof they're related, but a reminder that not all employees are there to contribute directly to UX. Administrative duties and other company responsibilities are likely suffering due to these cuts. This may also be playing some part with twitters tiff with advertisers.
Beyond that it would appear that a not insignificant number of these cuts were outside the US. Indian offices had 90% of their staff cut, for example. I have no idea what the implications are for this, but many of the impacts on twitter UX might just not be happening in Scott’s region.
>> “Although stories from this winter claimed that Twitter Blue was a dud, anecdotally I’ve been seeing lots more people using it lately. This could provide X with a revenue source independent of advertising and make them well-placed to survive any future chatbotpocalypse.”
As for this I am fairly certain that's just Scotts twitter bubble. Twitter blue is still relentlessly mocked across much of twitter. An amount of accounts that find value in the benefits for outreach for content/business/whatever do exist - reluctant or otherwise. But that's not a significant enough demographic to rely on at its current cost. And the benefits it offers do not warrant higher pricing. For the average user blue still just is not appealing. For the most part it's relegated to the afformentioned accounts that highly value the benefits, or Musk fanboys/tech adjacent enough accounts. Which is fine, but feeds back into the central conceit of this whole micro-discussion: Musks successes can outperform Musks failures. The pipeline from his PR ability to twitter blue sales and thus an alternate revenue stream is so direct it feels like a contrived hypothetical built almost exactly to explore this topic.
Thanks. It’s a truism that “other people don’t use Twitter the same way you do” (randomtweet.com used to be a great demonstration of this, but it doesn’t seem to work anymore) and many people said they were noticing much worse user experience.
The twitter bubble you inhabit definitely makes a big difference to your perception here! The vast majority of users in my niche of twitter have left, nobody has twitter blue; the site is just a fairly boring wasteland for me now, punctuated with the odd promoted Elon tweet which I have no interest in. I went from a decent amount of daily use to only checking it once or twice a week.
For example, I’m surprised to hear this! I thought there was a week or two when everyone threatened to switch to Mastodon, then found they didn’t like Mastodon and went back? So where did everyone go? Was it Mastodon after all? Facebook Threads? Blue Sky? Or did they all start learning to paint and spending time with their friends and families?
Dan Lucraft writes:
Community Notes existed before Musk and I don’t think he’s improved it in any meaningful way. He did change its name from Birdwatch . . . It was in beta when he bought it. Or it had just been rolled out or was in the process of being rolled out.
This would be a big update for me if true: CN became much better a few months ago, and I was prepared to accept this as part of the Musk-can-accomplish-amazing-things narrative. But if it was just a long-term project finally leaving beta, that would take away his biggest Twitter accomplishment.
Matt S writes:
To me the worst thing he's done at Twitter/X is elevate the blue checks to the top of the reply stream. Reading the top replies was my favorite part of the site. Now you have to wade through dozens, or hundreds of posts of nonsense to get to actual good posts. And since no one does that anymore, good posts don't get much replies or likes anyway.
My time spent on Twitter has dropped to almost zero for this reason. But on the flip side, my productivity and outlook on life have improved significantly. So maybe I should thank him.
Bob Frank (blog) writes:
>> “He took over Twitter because he was addicted to Twitter, got a seat on the board, and then the other board members said he had to behave and he didn’t want to.”
You write further down about people making the mistake of misunderstanding Musk because they dismiss what he says and don't take it literally. Well, what did Musk literally say about taking over Twitter? He said that people pushing toxic identity politics caused one of his children to become estranged from him, and Twitter was so deep in the toxic identity politics camp that he was literally not allowed to talk about the harms it was causing, so he took over in order to clean up the cesspool that Twitter had become.
This is a good point. I’m confused because it definitely looks like he tried to get a board seat, and only took it over after that was going to be unstable, then tried pretty hard not to take it over in a way that suggested he wasn’t that committed.
Here is an article claiming that his ex-wife Tallulah Riley begged him to “buy Twitter and then delete it” because: “America is going INSANE . . . [Twitter] is very easy to exploit and is being used by radicals for social engineering on a massive scale. And this shit is infecting the world. Please do do something to fight woke-ism. I will do anything to help!”
Anyone who has this good a relationship with their ex-wife can’t be all bad. But also, based.
I do disagree with your stance on Twitter/X on a couple of fronts. In terms of Twitter Blue, I don't see it as successful or on the path to success because it is kind of a half-measure. Twitter should either be primarily subscription model that basically operates on "FOMO" for its addicted users or rely strictly on advertising and be the town square. Twitter Blue is in this weird in-between where there are benefits to membership but its not needed to use Twitter, and so it becomes just a poor enough experience for non-subscribers that I think over time they will start losing users as the site becomes less useful (Like seeing subscriber tweets and comments before the people you actually want to read).
The second issue I have is with the idea advertisers need Twitter instead of vice versa. There are plenty of social media options for advertisers to go to, I just don't believe Twitter is essential there. Especially when Twitter isn't really associated with buying things as a consumer. No one shops on Twitter. I can see this actually being a major problem if advertisers stay away and realize their bottom line isn't actually hurting, so might as well just make this decision permanent since its all downside being in the Elon Musk business from a PR perspective (As your own review noted in terms of how he handles PR and drives folks nuts).
I wrote a piece for Foriegn Policy about why the vision of an 'Everything App' in the US market is essentially impossible. WeChat was created in a very specific Chinese context that simply doesn't translate to the US
Another example about getting himself into a lucky position is buying Twitter and now being able to use that data for an LLM project, he didn't anticipate that but considered it a nice bonus.
AI experts, is this a big deal? Can other AI teams not access Twitter data through the public web? Is it a substantial amount of text compared to other corpuses? Is the structure (280-character blurbs written by morons) a limiting factor? Or is this a genuine treasure?
7: Comments About Musk’s Mars Plan
Unirt pushed back against my claim that colonizing Mars wasn’t useful, saying it was a good way to avoid extinction if Earth got hit by an asteroid. I wrote:
You can solve the asteroid threat more easily by tunneling underground on Earth and building a colony beneath the surface.
I don't even know if it has to be underground - what if you have self-sufficient domes in a few different places, so that if the asteroid strikes (eg) America, you've got a dome in Russia which isn't destroyed by the shockwave itself, and is able to sustain itself agriculturally for a few decades until you can go outside again?
Also, the chance of an asteroid strike is about 1/1 million in the next 100 years, so bringing a Mars colony forward by 100 years doesn't get you a lot of x-risk reduction.
The major problem from an asteroidal impact on the order of Chicxulub is the dust and soot in the upper atmosphere cutting off enough sunlight to make growing plants impossible for as long as a decade. No protective dome or underground bunker will make the upper atmosphere transparent enough to grow plants. And even if you've got enough preserved food, vitamin C has lousy shelf life. Hope you've got enough light bulbs and a solid enough supply of electricity to grow enough cabbage to make enough sauerkraut to ward off scurvy.
Not that it's logistically easier to maintain agriculture on Mars than set up survival bunkers with lots of grow lights around a nuclear power plant or something, but any major settlement on Mars will likely, simply as a matter of cost reduction, have substantial local greenhouses taking advantage of the natural sunlight on a reasonable day-night cycle, plus reasonable local supply of CHON elements. If Earth's atmosphere is rendered opaque, a large Mars colony might manage to keep humanity going.
Of course, that requires there to be some reason there's major settlement on Mars to begin with.
I refuse to believe that going to Mars isn’t 100x more expensive than figuring out ways to solve these problems on Earth.
This is something I think Scott's review didn't emphasize enough about Musk and how to understand his persona. You can look at nearly everything Musk does up to the 2015 book with an eye toward building a colony on Mars. Something like SpaceX is easy to make the connection. Tesla and their solar subsidiary also make sense. But what about the others?
The Boring Company is exactly the kind of infrastructure project you'd need on Mars, where digging tunnels is basic infrastructure.
The same goes for the hyperloop. How do you travel long distances on the red planet? Are you going to have airplanes? There are no fossil fuels, and the atmosphere is thin. A few years back, Musk said he thinks it's technically possible to build electric planes with very long ranges (on Earth), but he's not interested in the project. Probably because it's not practical for Mars. Meanwhile, a hyperloop would be technically much easier to maintain in the Mars atmosphere, and would probably be better than planes for long-distance travel.
More recently, I think he has tempered his expectations on Mars. He talks about it less, and seems to be doing more that isn't obsessively focused on a Martian colony, like buying Twitter and focusing on AI x-risk. However, I wouldn't be surprised to see him make a new push in a few years to build out some kind of space exploration company that everyone thinks will fail.
And this is the part where I think understanding Musk really takes shape. Elon doesn't seem to be driven by a desire to build companies so he can make more money than anyone else on the planet. I'm sure he enjoys that, based on some comments he has made. The book mentions his intense partying, but I don't get the sense that he's working 100+ hour weeks for the money.
Sometimes he doesn't even seem to be driven by whether the companies nominally succeed. Yes, he cares, but when asked what odds he gave SpaceX and Tesla of succeeding, he said about 10%. The obvious follow-up question is "Why do them?" Why sink his ENTIRE fortune into companies that combined don't look like a good bet for making money? Because, he explained, even if they failed they would accelerate rocket ship development and adoption of electric cars. And to Elon that was worth the price.
Maybe Elon's "secret sauce" is that he's not laser-focused on profitability. He doesn't measure success in EBITDA, stock price, or any of the other metrics people use to gauge success. I'm sure he still stresses over these metrics, but users gained/lost doesn't cause him to change course or abandon a new project. He simply says, "I guess I'll have to adjust my expectations about market share" and pushes onward in his quest to make the Everything App. You think reusable rockets are a waste of money on an impossible dream? Elon doesn't care, because you're not going to get routine space flight without them and his goal is routine space flight, not making a bunch of money building rockets. So he's going to make the rockets reusable if it costs him the company. Because to Elon, it's the idea that drives him.
I don’t know about Hyperloop. The book made it sound like he originally proposed it because he was angry about California’s less-ambitious high-speed rail plans and didn’t take it too seriously. It was only after everyone else took it seriously that he gave it a second thought.
Likewise, Musk was an early investor in Tesla because the founders approached him, he thought it was a great idea, and he’d been interested in electric cars since childhood. I don’t think there’s a lot of room for him to have planned ahead how it would synergize with Mars, even though it does.
I agree with the broader point about idealism.
It is completely obvious and heavily signaled that Mars is about liberty / independence as far as Musk is concerned.
"Starlink's terms of service include a Mars clause: Users must agree that Mars is a free planet unbound by the authority or sovereignty of any Earth-bound government."
Zubrin's Mars Society was an early influence on Musk, and Zubrin's whole thing is that Mars is a new America, free from the old world's stultifying influence. "No EPA on Mars is one of the major reasons we have to go there" among other quotes.
It's somewhat annoying that people act like the Mars thing is out of pocket. It's the only planet besides Earth where you can sit on the surface and get as much CO2 as you want. It has readily available water ice. You can launch a single stage rocket from Mars to anywhere in the solar system because of the shallow gravity well and thin atmosphere. It is the only planet where independent survival is possible with near-term tech. Living on another planet can't be compared to Antarctica.
This review/book seems to presume that he has random obsessions and just happened to make electric cars and rockets -- it's pretty clear that rocketry needed major improvements and that those improvements were technically possible, same with clean energy/electric cars. Those were also at the burning core of America's national interest. This is a man who loves sex, video games, and porn - an accomplished womanizer who jets around including with Hollywood actresses. He's ravenous for all that life has to offer and takes a big bite out of everything. Yet he'll still buckle down and work day and night on what he thinks is important -- sacrificing an AAA class lifestyle long periods of time. How many people have "made it" and kept swinging in that manner? None of the other major tech people seem to have that fight in them - they shoot their shot and then move on to their foundation. Where's Jeff Bezos? Where are Larry and Sergei? Where's Bill Gates?
Yet more than any of the other tech billionaires, who "hit it and quit it," or who knock one thing out of the park and whiff thereafter, he has gone back to bat time and time again, and each time for something incredibly important. Yet despite being the most admirable and principled tech magnate in this most important of regards, he has by far the worst reputation because, essentially, of his vibes.
Maybe his irreverence for norms should reflect badly on the norms, not on him.
I agree with this except for the “where’s Bill Gates?” question - Gates is saving millions of lives running one of the greatest charitable foundations in history. Even if you generally agree with the market > charity hypothesis, the Gates Foundation might be the one exception!
Ethics Gradient writes:
I really have to give Musk credit for just embodying the kind of arc-of-history techno-optimism that seems to have been a hallmark of the space age and just abandoned for blinkered, ultimately transient concerns that are just less inspiring and less long-term relevant than the kind of challenges Musk has taken off.
At a certain level, "Why should we go to Mars?" has to be a question that answers itself, and I think Musk is the primary counterweight against various forces that would pose it. Pose it in a manner that makes short-term sense but at the cost of any sort of long-term vision for human advancement. Perpetual local optimizations can generate a tremendous amount of value but it is not a total replacement for an overarching teleology. Wisdom is slave to the passions -- let us have our damn passions back.
Telos is humanity reaching out into the uncaring universe and wresting meaning from the void by force of will. It's worth going to Mars *because it's fucking going to Mars.*
8: Comments Comparing Musk To Other Famous Figures
Bill Benzon writes:
Musk's hands-on learn-how-it's-done practice reminds me a bit of Walt Disney, of all people. Do you know how Disneyland came about? During WWII Walt became interested in (obsessed with) model railroads. So he went into the company's machine shop and learned how to use the equipment to build his own models from scratch. Then, after the war, he had the idea that his employees needed a park where they could relax with their families. That much I learned from two biographies, a thick tome by Neal Gabler, and a somewhat more sympathatic one by Mike Barrier (and, I believe, a bit deeper). So, you merge the idea of a park with the skill of building model trains and out comes Disneyland, the world's first theme park and, some have said, a masterpiece of urban planning.
Disney, too, has been a controversial figure, very.
J. Ott adds to the Disney comparison:
Walt Disney has some parallels - was a technology innovator and pushed workers beyond what they imagined their creative limits. He also had his very sensible brother Roy handling the finance side. (The company was called Disney Bros. until Walt decided it sounded better with just his name.) The usual pattern was that Walt would push the company the point of bankruptcy on a dream project, Roy would call in favors to get a bunch of bankers in a room to hear about it, then Walt’s incredible pitching ability would convince the skeptical bankers it was worth keeping them solvent a little longer. When all else failed they would make a princess movie. What I take away from the Walt biographies is that Roy was working hard behind the scenes to keep everything afloat and that with him (and genius Ub Iwerks), Walt would have flamed out many times.
Elon started his first startup, Zip2, together with his brother Kimball. But I don’t get the impression Kimball was much of a moderating influence on Musk; if anything, they were too similar. Here’s what Vance had to say about their relationship:
[Early investor Greg Kouri] used to referee fistfights between Elon and Kimbal, in the middle of the office. “I don’t get in fights with anyone else, but Elon and I don’t have the ability to reconcile a vision other than our own,” Kimbal said. During a particularly nasty scrap over a business decision, Elon ripped some skin off his fist and had to go get a tetanus shot. Kouri put an end to the fights after that.
Eric Zhang writes:
Napoleon Bonaparte reminds me of this. Or rather, reading about Napoleon reminded me of Elon Musk - particularly the terrifying intensity and emphasis on speed, the micromanaging, the ability rapidly to suck information out of people's skulls via conversation, the incredibly bold bets.
James McSweeney agrees:
Elon's personality appears amusingly similar to that of Napoleon (at least, as related in Andrew Roberts' 'Napoleon the Great').
Both boast/boasted near-perfect recall (Napoleon could recognise and name common soldiers he'd met two decades prior), a widely remarked-upon ability to rapidly master the complex details of processes at every level of thier operations (+esoteric topics of interest), a tendency towards obsessive micromanagerial interventions, and reputation for meeting timelines conventional wisdom deemed impossible, though a combination of belligerence and highly motivated employees. Both worked their way out from initial training in maths/engineering, were/are obsessed with the frontiers of technology, and think/thought in terms of arcs of history. The megalomaniac box also probably gets a tick in both cases.
Is Napoleon what happens when would-be tech bros lack silicon?
Key differences between the two are that Napoleon had a habit of being (often unjustifiably) trusting of long-time colleagues, built rapport with his low-level employees, and was exceedingly charismatic.
After decades of risk-taking, Napoleon's luck eventually ran out. It will be interesting to see if Musk meets his own proverbial early Russian winter.
Schroden Katze writes:
Description of Elon really reminds description of Stalin from Mikoyan diaries
Very alike idea of ruthlessly appointing and firing people, same treating people like cogs, same very deep micromanagement where he both knew ton about ton and also faked a lot of it
He even did that very aggressive examining of his subordinates that for many looked like brutal test, but actually it was him figuring out
One anecdotal story tells how Stalin that Germans use electric melting of steel, so he drove right to home of minister responsible for the newest steel factory and accused him of sabotage for using coal
For the next hour a minister afraid of his life was proving and explaining all details of steel industry and why on certain mill the coal was preferable
The same was near-endless work stamina and also charm, for some reason people believed he was a great person no matter what Stalin did to/with them
I love Tolkien’s work, and, oddly, when I was reading this, my mind turned to Feanor in The Silmarillion.
Feanor was the most gifted of the elves. His was unbelievably brilliant, talented, and determined, and “his spirit burned like a bright flame.” He would work obsessively hard and he achieved many great things, including making the three Silmarils, the magnificent jewels after which the story is named.
He was also arrogant and quick to take offense, and he made some catastrophic errors in judgment that cost both him and his people dearly. He was very charismatic when getting people to follow him, but had zero kindness or understanding of others.
He had seven children, so not as many as Musk but still a large number, and he and his wife became “estranged “ due to his bad choices. Bear in mind that this is in a culture where divorce didn’t exist and true love was forever, so becoming estranged from your spouse was a really huge deal and a sign of something being deeply broken and wrong.
As Scott would say, TINACBNIAC.
9: Other Comments
Moon Moth writes:
If you look at all 11 of [Elon]’s children's names, all but the 3 with Grimes have fairly normal names, at least for American kids these days.
The list is (with Grimes children in bold):
Xavier (later changed name to Vivian)
X Æ A-Xii
Exa Dark Sideræl
All the non-Grimes children have pretty normal names, at least by Bay Area standards (some of my friends’ kids’ names are at least as weird). So is Grimes behind the weird names, or are the other women better at reining in Elon’s trollish tendencies? I find it hard to believe the kid named “X” isn’t Elon’s fault, so good work by the non-Grimes mothers putting their foot down.
One big thing I think you missed, which is connected to persistence that you mentioned but still a wholly separate method, is iteration. Everything is iterated upon until it works and after it works, its iterated upon even more. I "ctrl+f" iteration and didn't find the word mentioned once. It is basically taking software methods into hardware, and it isn't simply about changing things like you say, but iteratively improving.
Musk walks through it here:
One of his employees walks through it here (the timestamp is for a section where you can see the steps, but the explanation starts before):
FractalCycle gives the serious EA perspective:
Feels weird talking about Musk, since his biggest impacts are fuzzier ones on x-risk (cofounding OpenAI and also the Ukraine Starlink non-activation event). AI risk and global geopolitical/nuclear risk. So far, what he's done in those areas is questionable at best and unusually terrible at worst.
Taking near-term extinction risk seriously, even getting to Mars wouldn't necessarily outweigh nudging the AGI field in a more dangerous direction (i.e. if OpenAI has contributed more to capabilities than alignment, or if X.ai does anything big).
IMHO these are the 3 things (X.ai, openai, and Ukraine) that matter most about Musk, and so far he seems net negative. The other massive things are rounding errors in the face of that, yet get more attention. (The extreme case: Twitter/X is a rounding error *on those other rounding errors*, and ofc that gets discussed 1000x more than everything else.)
One other thing that doesn't seem to fit with this review, but that I'm not sure how to think about. Musk famously gets into a lot of fights with the US government, for example around SEC enforcement and the covid response. But we don't see him criticize China much -- maybe never publicly since covid? From a business perspective that seems smart: the Chinese government has a lot of ways to retaliate against Musk through Tesla, and it seems believable that they would if he criticized them publicly too much. But the review paints a picture of a guy who couldn't maintain that amount of message discipline on subjects he was really passionate about.
Maybe the fights with USG are more tactical than deeply-felt? Maybe China is just intimidating in a way that America isn't? I don't have a conclusion here, just a vague sense that I'm missing something.
Yeah, I think 1) he isn't politically coherent 2) he has the same 'mostly focuses on things that are salient to him instead of important' problem everyone has 3) he's selfish and knows Xi will ban Tesla the moment he takes on China.
Steeven (blog) writes:
FWIW, I have a few friends who work at the gigafactory and they have a picture of Elon in their front room as a saint and call him daddy. They also work 6-7 days a week.
He included a link to the picture, which is:
I admit I’m a sucker for this style of art; see also here.
My strongest update was shifting my estimate of Musk’s success at Twitter downward based on other people’s descriptions of worse experiences with the site, the “maintaining an alliance with the elite blob is a key part of Twitter’s business which it’s legitimate to spend 80% of its workforce on” argument, and the claim (still not proven, but plausible) that Community Notes wasn’t really a Musk project.
It’s nice to have Musk’s SAT scores, but they were about what I predicted so this wasn’t an update.
I’m more willing to consider the possibility that Musk might have bipolar by some definition, but still think on balance probably not.
The claim that the Boring Company’s key insight was that you can build tunnels differently with electric cars - was interesting, and I hadn’t heard before. But I’m not sure the Boring Company is interesting enough for this to matter.