astralcodexten.com
Scott Alexander’s review of a 2015 biography of Elon Musk. Elon Musk, to me, is one of the world’s most confusing people. He’s simultaneously both one of the smartest people in the world, creating billions of dollars of value in companies like Tesla and SpaceX, and one of the dumbest, in burning billions on Twitter. Scott’s review I think is a good explanation of what’s up with Musk.
Jump in the discussion.
No email address required.
Notes -
On twitter I have seen the likes of Cernovich and others claim Elon is among the smartest people on earth, if not the smartest. He's smart, no doubt, but to be among the smartest would put him in the company of Fields Medalists, math Olympiad and Putnam winners, math and physics doctorates, etc. Possible but very improbable, imho. He's smart but not that smart. He benefited from luck and good business savvy and intuition about markets, all necessary ingredients for business success. The Twitter Blue verification system did not stop spam, as I and others correctly predicted (when scammers can make $1000s in a day , how is $8 a deterrent). Maybe he was not smart enough to see that, or he did not care.
What do we mean by smart at this level?
I recall some Yudkowsky post talking about how excessive fixation on the definition of intelligence was unhelpful. If it runs away and makes a Dyson Swarm, then it's intelligent regardless of whether its conscious or not... If Musk can show up and start a bunch of multi-billion dollar technology companies, then he must be very smart. Intelligence is about making things happen, not solving tests. Who cares about maths Olympiad or Putnam winners? Are they as smart as Elon? I vote no, it doesn't matter if they are better at advanced mathematics or certain kinds of rareified, abstract problems. Who cares? Elon solved a bunch of problems in the real world, where it matters. Even if it's just managerial talent, that alone puts him far ahead of nearly all physicists.
Why try to distinguish business savvy, market intuition, management, perseverance and determination from intelligence?
If there is a tallest man in the word, it stands to reason there must be a smartest, or at least some people who are candidates for being the smartest. Is it Elon? probably not. Does it matter? Given his success, not at all . But I am not the one making that claim he is the smartest. Elon made a lot of money and did a lot of cool things, , but so did Google founders, Bezos, Gates, and Zuck...they are self-made (or had some parent's help ) and have very valuable companies. Elon founded many companies, but these are not statistically independent events. if you limit your pool to people in tech who founded billion-dollar companies, you still end up with a lot of people who are all pretty smart, but is Elon the smartest? Statistically speaking, likely not.
he may have solved a coordination problem, but did he code the AI for a Tesla self-driving car or does he understand the code as well as someone who coded it? probably not
Being the smartest is more like being the strongest than the tallest.
More options
Context Copy link
Suppose the guy who wrote the AI for the Tesla self-driving car is better at writing code than Elon, (I note that Scott describes Elon as stepping in to do the job of anyone in his team if he wasn't satisfied with their work). Is this coder a billionaire? Certainly not, he hasn't succeeded like Elon has. If he's so smart, why doesn't he go off, start his own company and make a tonne of money like Elon did? He likely lacks the all-important management skills and risk tolerance needed for great deeds.
The ultimate test is actually succeeding in the real world, not writing great code. If your code and ideas are truly fantastically great, then you can make billions of dollars, become world famous like Satoshi purely on the merit of your product. Nameless Tesla AI coder isn't in that league because almost nobody is.
Furthermore, management skills are more useful than technical skills. Institutions aren't merely the sum of their employees, NASA isn't struggling to achieve 1960s-tier goals because their engineers are terrible. If they sacked their engineers they wouldn't do any better. It's the culture, the policies, the structure of the organization that's decided/tolerated at the top. That creates the power of a company or institutions. Elon creates the conditions for the Tesla/SpaceX engineer to be so productive, just like Napoleon organized the French army such that its soldiers and generals fought harder and better.
I'm not prepared to say that Elon is smarter than Gates or whoever else but he is very smart, surely one of the smartest men alive, which is your original point of contention.
yeah but you can teach someone management. but not everyone can learn differential equations well. the difference between going to the moon or not is technical skills.
I disagree. Elite managers like Napoleon or Elon have something else that distinguishes them from normal people. Just like with differential equations, some people are never going to be even mildly charismatic and just aren't leadership material. Surely you've seen them, they don't want to talk, they fumble social things...
Just as there are mathematical and social dunces, there are mathematical prodigies and management prodigies. Elon had this thing where he could quickly judge whether someone was really competent in a few minutes and so he hires the best people and motivates them to give it their all. Napoleon could remember the names of his soldiers even after many years and form really close relationships with them, he had huge charismatic power. A bunch of people wrote about how Hitler would just verbally monster people, bully them into going along with him - he had a special power that entrances and frightens people even long after his death. This stuff can't be learnt at Wharton Business School.
Without really good management, the engineers can't reach peak performance, teams get bogged down with delays and overruns and waste. Otherwise, NASA would easily be able to go back to the Moon, it would be a trivial task! Engineers haven't gotten stupider, technology hasn't deteriorated, it must be management that's the missing link.
Or take the Boeing MCAS system that tried to turn 737's into dive-bombers. Billion-dollar companies like Boeing have access to engineers who can make guidance systems work. It's not that hard. They did it before. But because of their lax management, cost-cutting and parcel-passing, hundreds of people died:
yes becase the math and engineering was already figured out. management is important, as I agree with you there, but engineering is what stands in the way of something being possible or not, even with the best managers the world. if the equations cannot be solved, then it is literally impossible to get to the moon. not impossible in an abstract sense, but literally cannot be done.
That's the difference between necessary and sufficient. Yet sufficient can be way more important than neccessary.
Copper is necessary to make phones. If you have no copper, you can have no Iphone, categorically. But it's in no way even remotely sufficient. The Trojans had copper! >5% of the work is in raw material extraction. 95%+ of the value is in stuff like refining, design, chip manufacturing, software... Iphones need copper but they're not about copper like a spear might be.
Maybe in 1750 it really was about finding an engineer, or better an engineer who knew what he was doing. But today there are loads of engineers who can do pretty straightforward mathematics, the real substance of all our problems is in management, policy, quality-control, supply chains, efficiency...
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link