world destroyar

seeking sensorimotor expansion

[world destroyar @worlddestroyar]

We are machine-makers. When we have a problem, we make a machine to solve it. When we have many adjacent problems, we make a machine that makes the machine that solves each. And so on. There is nothing more to life.

[world destroyar @worlddestroyar]

@natural_hazard https://t.co/YhINvSfiG2

[world destroyar @worlddestroyar]

The bottleneck is amplifying the influence of sane thinking on sheep behavior in the near future and language models have come in clutch. Or perhaps not. We'll see.

[world destroyar @worlddestroyar]

@natural_hazard https://t.co/0GRpsRvEMZ

[world destroyar @worlddestroyar]

@sbaroncohen https://t.co/73BfwKIuHI

[world destroyar @worlddestroyar]

@davidbessis I do value it and I'd be happy to fund it privately one day. I don't understand how you can justify shaping public opinion to trick random people into funding it. You called me a loser, now I'll call you a communist. You are French, after all.

[world destroyar @worlddestroyar]

@davidbessis Yes, of course. First by me, then tutors.

Could you please make your tweets denser? It's not very useful to go back and forth like this. What are you getting at?

[world destroyar @worlddestroyar]

@davidbessis Definitely not because I remember school (private, supposedly one of the best).

[world destroyar @worlddestroyar]

@davidbessis not yet

[world destroyar @worlddestroyar]

I haven't since I was 17. They do get some help from me though.

The kind of public opinion profitable businesses depend on (market prices) is very different from the kind mathematicians depend on (projecting an air of superior knowledge and ability). I enjoy being around mathematicians but I do wonder who pays their salaries and why.

[world destroyar @worlddestroyar]

@davidbessis I meant mathematicians. I know you made money by selling ads

[Cameron Schwartz @nyoomtm]

Liftoff of Starship V3, from the dunes right outside the pad.

This is the most insane shockwave action I have ever seen on video. Absolutely mad.

📽️ Me for @WeAreSpaceScout

[world destroyar @worlddestroyar]

@wolftivy @prerat Irrationalist bullshit. So weak, Wolf Tivy

[world destroyar @worlddestroyar]

@CestrianInc on the other hand https://t.co/mhdrIoatiz

[world destroyar @worlddestroyar]

@0xthestudents Let me see more of Landau-Taylor instead of this pls

[world destroyar @worlddestroyar]

@tailormadecloak @ultimape nvm https://t.co/gFcirvnlcR

[world destroyar @worlddestroyar]

@ultimape @tailormadecloak btw, I'd really love to get a copy of your export to analyze it

[world destroyar @worlddestroyar]

@HiFromMichaelV huh https://t.co/Iz7fyKGfYI

[world destroyar @worlddestroyar]

Would language models based on diffusion with a more complete covariance structure be better at humor than autoregressive models? Can AR models trained in reverse direction reveal something about humor? A joke's structure seems to be: ABC likely, BC less likely than BD where C is the punchline.

[world destroyar @worlddestroyar]

So, you're looking for speech patterns that would have required an LLM to have some pre-thoughts. The text that gains the highest surprisal multiple with the least surprising pre-thought. @HiFromMichaelV

[world destroyar @worlddestroyar]

@aman_unsiloed You say it's "simple per-page pricing" but charge a subscription fee? Why?

[world destroyar @worlddestroyar]

@vlada_mc Fat coupled with carbs seems bad, so a fat-heavy zero-carb and a carb-heavy zero-fat diet would both be ok.