Wei Dai

Mail sign emoji.

Wei Dai, look at this from the other side, of someone trying to do long horizon strategy.

You spend most of your time having low status for being obsessed with something nobody else cares about, and then can only get points for being early, not for the details of your strategies slash ideas, which we still can't evaluate.

Did the early EAs slash rationalists obsessed with AI safety make the situation better or worse, and how to divide up the credit among them? It's genuinely hard to tell, even now.

He's an American citizen.

And Wei Dai, who graduated from the George Washington University.

And I guess if you look at it, I think Adam Back is British, so he's a bad example.

Also, Nick Sabo is not mentioned in the white paper, but we know that he had the closest idea to Bitcoin with Bitgold.

Because it's hard to evaluate this work, street lighting steers away from it.

It's also arguably an illegible problem in Weidai's usage.

Prediction is very hard, but also very useful.

Actually funding people to do it and publish it full-time might be better than everyone just doing little pieces in their spare time.

It's like a cognitive dissonance to live in these huge structures, but at the same time have this brain that hasn't adapt-adapted yet.

Because of that, I think that the cypherpunks since the '80s, starting with, uh, Chum and, and then with, uh, uh, Tim, uh, uh, Tim May, may he rest in peace, and, and Nick Szabo and Wei Dai and, and Hal Finney and all the other cypherpunks, they, they maybe rationalized this, um, and, and, and they were very strong advocates of individuality, you know? And, uh, and they thought correctly, I think, that, uh, cryptography would-- they could build, build with cryptography these online, uh, environments where we can really create, uh, individuality and do, do the reverse.

In-instead of blurring the individual, to blur the structures above, you know, the top-down structures.

And I think that, uh, in that sense, uh, Bitcoin is the response.

Um, yeah, in my case, I, I, I don't...

I can't claim that I understand, but I, I can claim that what I understand, uh, uh, from not only from Nick Szabo but from Tim May, um, with whom I had a lot of interaction in the last, uh, three years.

Um, Elaine Ou, another thinker, um, Hal Finney, uh, Wei Dai.

Um, when you read all these things from, from cypherpunks, uh, it's incredible.

On the other side, he critiqued people who claimed that aligning prosaic AGI is probably infeasible.

As Wei Dai pointed out in response to my previous post, Miri used this claim to justify trying to build agent foundations-based AGI themselves.

So Paul seems to have been trying to steer a path between two opposing alignment strategies which both prescribed building AGI yourself.

An admirable intention, if so.