Hacker News Reader

I'm not sure what's unclear about what he's saying.

a) Harvard is a pure signaling mechanism that doesn't provide education as much as exclusionary credentials.

b) AI/Big data is a way to solve the economic calculation problem that was once seen as one of the main theoretical hurdles against communism. That a country like China favors AI and punishes crypto tells you something (that it's a communist dictatorship).

It's pretty straightforward.

I also liked the fact that he points out he'd like to see innovation not just "in some iPhone app" but more broadly in different fields.

I think this is a good point, there isn't enough going on in other areas, like he said people ride the 100 year old subway distracted by their iPhones not realizing the world around them is unchanging mostly.

It's true in some ways, who will build or even demand the better faster subways of the future if 99% of the people don't even realize they're on the subway.

Yup. To be fair, that's a point he's been making for ~10 years about the world of atoms and world of bits dichotomy.

I think it's also, in part, subtly caused by our narrow understanding of the word "technology" to be "software made in Silicon Valley" as opposed to a broader "tool that serves a human purpose."

If we broadened that understanding, we could see engineers applying their skills to problems that would be just as lucrative, but could be more legitimately called "innovative" than working on the umpteenth marginal "tech" app.

Well, but China doesn't use AI to centrally plan the details of its economy. Corporations, prices, and competition are undisputed features of the economy there. Whether you want to call it communist or not, the computational problem is still as intractable as ever. There's a difference between monitoring and regulating society with computers and solving the classical economic calculation problem.

And Harvard is a way for elites to network, but that's not merely arbitrary signalling - it depends on how those elites are selected.

Sure. Thiel is proposing a model. All models are wrong, some are useful, and so on. His models happen to be interesting to people who like to think in more contrarian ways. That doesn't make them right, and we can debate their merit, of course. But it hardly makes them impossible to understand.

That China hasn't solved the economic calculation problem yet is plain to see, and I don't think that's quite what Thiel is saying. He's talking about the principles and the mindset behind the support for AI, as well as its ramifications in terms of the type of society it can produce asymptotically.

No one said Harvard's signalling was arbitrary. It's a very clear signal. Thiel's point is that that signal has nothing to do with "higher education".

"well as its ramifications in terms of the type of society it can produce asymptotically."

My assertion is that it can't though. That's because the complexity of even a fairly small entity makes the number of the particles in the universe look infinitesimal.

https://en.wikipedia.org/wiki/Busy_Beaver_game

"In 2016, Adam Yedidia and Scott Aaronson obtained the first (explicit) upper bound on the minimum n for which Σ(n) is unprovable in ZFC. To do so they constructed a 7910-state[2] Turing machine whose behavior cannot be proven based on the usual axioms of set theory (Zermelo–Fraenkel set theory with the axiom of choice), under reasonable consistency hypotheses (stationary Ramsey property).[3][4] This was later reduced to 1919 states, with the dependency on the stationary Ramsey property eliminated."

To me, granted I have only a vague idea of the meaning of the above, that's saying that "solving the economic calculation problem" (society having more than 1919 moving parts) is not just beyond physical limits of the amount of matter in the universe that could be used for computation. It's not just beyond the physical laws that we think we know, like a FTL spaceship would be. It's beyond the most fundamental math we have. To me, that's like a third order of infinite impossibility.

And as far as Harvard goes, how can the interaction between students and faculty have nothing to do with higher education? What else is higher education? In every school, people learn from each other, and they interact more than with people at other schools.

Rightly or not, Thiel is conflating communism and totalitarianism- China is indisputably a leader in using AI to perpetuate the latter.

I thought I was making it clear I don't care whether China is labeled communist or totalitarian or whatever, nor whether the author labels it so.

My point was that they are not using AI to do the sort of central planning that earlier communists failed at and that economists eventually decided was infeasible. And I don't think there's any prospect of that happening in the forseeable future either.

Maybe not so elegantly put, but there's no question that China has embraced AI as a tool for spying on its own people. For whatever reason, spying seems to be a consistent Orwellian feature of communism. It was a central feature of the Soviet Union and China just seems to be taking the franchise to new heights with its approach to AI.

I don't know what to say about the Harvard quote, but with the recent admissions scandal implicating that rich people are buying their way into premium universities and into higher SAT scores, there's some nugget of truth to criticizing the current situation.

I wonder if universities aren't suffering Goodheart's Law: "Once a measure becomes a target, it ceases to be a useful measure." That is to say, once we measured that a university degree lead to higher incomes, better life outcomes, etc. and then we started pressuring everyone to get a four year degree...did it cease to be a good measure for quality employees, creative thinking, problem solving skills, etc?

I think there's something to that. When everyone is obsessed with getting the degree, and they're not worried about the practical application of their knowledge, they are doing the world a disservice. And we're all suffering for it, through higher debt and inflated expectations.

I'm not sure America is as different from China as you seem to imply here. You don't manage to incarcerate more individuals than "an authoritarian communist dictatorship" four times your size without massively scaling the state's ability to monitor citizens.

20% of the world's prisoners are in America. Is spying a consistent Orwellian feature of American capitalism?

Isn't this the guy who founded Palantir?

Quoting him and restating the point isn't an argument. AI fits very neatly into China's policy goals in a way that it doesn't for liberal politics.

We don't know how the improvements in AI are going to play out - it is very easy to see how it could turn out that these are the missing technologies to make a totalitarian state the most productive equilibrium, or communism, or anything else.

> And then, it is literally communist because China loves AI; it hates crypto. And so that, I think, tells you something.

China also loves cheap r/c cars, does that make them communist?

"Crypto"? The biggest Bitcoin miners are in China. That's because mining Bitcoin is "exporting", and a legit way to get yuan into dollars or euros. Buy mining hardware with yuan, mine, sell Bitcoin outside China. Not only is it legal, because it's exporting, it's encouraged and some breaks for creating an export business are available.

Exactly. Obvious factual errors on Thiel's part tells he's lost the plot

"Crypto" in context clearly means cryptography, not cryptocurrency.

Thanks, I was wondering if I should read it or not and this gives me some good tidbits to help decide.