the legibility bottleneck

 

 

What's the bottleneck that limits making innovative American companies? Why wasn't something like SpaceX made earlier? Why is new advanced manufacturing being done in Asian countries instead?

 

 

it's not capital

America now has more wealth than any nation that ever existed. Interest rates on Treasury bonds have recently been lower than inflation, and savings accounts at banks have had interest rates close to 0.

 

 

it's not engineers

American universities produce relatively few graduates in fields like material science, chemical engineering, and aerospace engineering, but there have still been more graduates than jobs. Many people with Bachelor's or even Master's degrees in those fields have had to go into other areas they're less interested in.

Of course, that doesn't mean companies say "it's easy to hire people". Have you ever seen stories about Americans in some field being particularly easy to hire? I've only ever seen stories about "shortages".

 

When there's more supply than demand, companies don't hire quickly - instead, they become more selective. Things you see include:

- only hiring people with some years of experience, not people from other fields or fresh graduates
- higher credential requirements: a Master's for what used to require a Bachelor's, and a PhD for what used to require a Master's
- not hiring people over 50
- several interview stages over a long time
- not hiring people with anything weird about their resume or career path

 

Companies hiring people on H1B visas doesn't mean there's a shortage, either. It's usually about cost, control, count, and/or caste.

- cost: Sometimes it's significantly cheaper to hire a H1B than an American citizen. Yes, companies are required to "try to hire an American" first, but it's not hard to pretend to do that. I, personally, was once told in a job interview that the interview was a sham being done so the company could bring in a H1B employee.
- control: People on a H1B visa often have to leave the country if they lose their job, and would have worse job prospects than an American citizen. That gives their employer much more power over them than normal. Management likes this.
- count: Managers will say things like "we have a team of 20 PhDs" to justify their budget and importance, as if PhDs are both fungible and a different species from people with other credentials. For this purpose, someone on a H1B visa with a PhD from a diploma mill is better than someone with a Master's from CalTech. Here's a little "secret" for those managers: a PhD involves studying something very specialized, and doesn't mean someone has a good understanding of that entire field.
- caste: A notable special case with H1Bs is Indian-Americans trying to hire people from the same caste and region, especially at tech companies.

 

Politicians usually hang out with journalists, CEOs, major investors, and other groups of people with little idea of what's actually going on at a low level, so it's pretty common to hear American politicians say something like "we just need more STEM graduates".

 

 

it's not ideas

If you go to investors, they don't want to hear about your idea.

If you become a grad student, your professor will probably be more interested in having you work on an existing project that got grant money than in you working on your own novel idea.

If you talk to CEOs or directors, they don't really want to hear about your brilliant idea.

Does that sound like an idea shortage?

 

 

the legibility bottleneck

Investors and executives rely on the university system to evaluate the technical competence of people. When technical knowledge is critical for something, they like to hire PhDs or professors, and get evaluation from other PhDs and professors. The problem is, that doesn't work.

Theranos had a Stanford professor on its board, and it was a scam. (It had a bunch of influential non-technical people, too.)

If you invest in whatever has a lot of papers published on it, you're gonna have a bad time.

If you try investing in technologies noted in MIT press releases, you're gonna have an even worse time.

 

 

So, what are investors supposed to do? What they're actually doing is: give up on technical due diligence and invest in whatever surface aspects they think will appeal to later investors.

Well, what alternative do they have? Learn about the technical details of a field starting from nothing, and use that to do their own evaluation? That sounds absurd, but that's exactly what Elon Musk does, and it's worked out reasonably well for him, in multiple disparate fields.

A common response to that is: "Elon Musk is a unique genius, and you can't expect other people to do something similar." That's not correct, but:

- People working for him will kiss his ass sometimes by overstating his understanding.
- If Elon isn't actually that unique, that makes CEOs and directors look bad.
- If it's possible to do better than relying on university credentials by doing things yourself, that makes universities look bad.

 

For those reasons, lots of powerful and skilled people have been incentivized to overstate Elon's abilities, but his actual ideas have been rather bad - not that many current American leaders are able to evaluate them, of course. Personally, I think his enthusiasm for working is more of an outlier than his technical understanding. CEOs that can't do the same thing for a single company and a single field are either just not particularly smart or very specialized in doing...something else.

Elon Musk is significant because he's at the bottleneck of innovative industry in the USA: the legibility of technical expertise to investors. He learns the technical basics of something, which is enough to fire some of the obvious fools. He's also known to people with money as someone with at least some technical expertise. That combination is the bottleneck, but not because Elon Musk is a unique kind of person. No, that's the bottleneck because America has no system to create people who fill that role. It used to use universities for that, but that responsibility has been abdicated.

 

 

I'll say it again: the state of US universities is bad enough that a smart and motivated but fairly ordinary person can do better than relying on university credentials by learning about a new technical field from scratch.

That's bad, but the situation is only going to get worse. The president of Stanford faked data. A bunch of Alzheimer's research was based on fake data. There's a lot more fake science out there, but people in labs are afraid to bring attention to it when they notice. Such people are now the leaders in place, in charge of choosing the next generation. Newer professors at US universities are more focused on metrics over substantive research than older ones. Graduate school has largely driven out people driven by curiosity and replaced them with people driven more by desire for credentials.

US universities are in decline, but the institutions that rely on them, like investment firms doing technical due diligence, don't seem to be adapting - beyond some of them giving up on doing what they used universities for.

 

 

 


back to index