jstanley

It's reassuring to know that even a supergenius's ChatGPT session is one sentence from the human followed by 3 pages of LLM output.

Were you using the Gemini image generator though? The article is about using LLMs to do drawing via tool calls.

I pasted the introduction into Pangram and it said 100% AI.

I too judge books by their covers.

I'm sure the track surface and shoes are important, but if "singlet" means the clothing he is wearing? It's really hard to believe that makes a material difference.

Trading on polymarket for example?

if the government fines oil companies for flaring off the natural gas, suddenly it's economically viable to capture it instead of wasting it.

If the government fines individuals for not digging holes in their back garden every day then suddenly it's economically viable to dig holes in your back garden every day, but it doesn't mean it's overall productive.

Agreed

there is less money remaining.

In what sense?

This may be related to the commonly-held fallacy of "cash on the sidelines". Cash is always on the sidelines. Cash is not created or destroyed by buying and selling stocks or bonds. Cash is simply handed from one party to another, but the cash has to be held by somebody.

Sure, try it. It's harder than you think. It's not just imports, it's the entire program.

you're suggesting there's some kind of information flow from the anticipated body of the script back up to the imports

Yes, I am suggesting this. I don't think it is possible to write programs without either anticipating what you're going to write down below before you get there, or else being able to go back and edit what you already wrote.

Of course agent harnesses allow the latter, but raw models without a harness can still do an exceptionally, superhumanly, good job of straight-line programming with no editing.

Infer imports from context, infer body from context + imports. All strictly causal.

Of course it's causal, that's kind of a reductive way to look at it.

Just infer the entire program from context and then type what you inferred.

You can't become a pro footballer just by wishing your leg wasn't broken, but you can pay close attention to the difference between pain and suffering, and acknowledge the pain without accepting any unnecessary suffering.

Pain is part of reality. Suffering comes from wishing reality was different to how it is.

It makes intuitive sense. How else could you write a 500-line script top-to-bottom with no backspace key and no arrow keys and get all the imports etc. right upfront?

It's because I live in the UK

Interestingly web.archive.org is blocked for me but TFA isn't.

The Tesla one talks about solar panels.

Really interesting list, thanks.

The link to the original for Scion Capital is a 404 FYI: https://billiondollarpdf.com/entry/burry-scion-subprime/ - the original seems to be included towards the end in the PDF at the top of https://www.michael-burry.com/scion-capital-michael-burrys-l... and is fascinating reading.

What does "strict receipt" mean? Every entry seems to be labelled "strict receipt".

EDIT: The Warren Buffett one is also a 404: https://billiondollarpdf.com/entry/superinvestors-graham-dod... - what's going on? Are people taking these PDFs down in response to high traffic from this site? Or you put these links up without checking them?

https://www.superinvesting.com/pdf/The-Superinvestors-of-Gra...

Conversely, https://www.lesswrong.com/posts/fzeoYhKoYPR3tDYFT/beware-iso...

I bet when hot and cold running water first came out you would have questioned whether people "needed" that as well.

Plainly it was not a "need" since people managed for thousands of years without it.

This is what progress looks like.

Isn't that the same hand?

Right but given that the agent only outputs text, and agents are perfectly happy writing python, the supposed benefit of Lisp is completely irrelevant here.

It's fine, if you don't want to read it you don't have to read it, nobody's forcing you.

Possibly, but people were worried about this with cloud hosting when it first came out, and it turned out to be a total nonissue.

One reason to want to upload the entire codebase is that it allows them to have the model inspect the codebase during "thinking" without going back to the client to do real tool calls.

It's not a really great reason, because what's the downside of going back to the client? But that's the best reason I can think of.

In what way?

The standard argument is:

Assume H(P,i) to be a program that tells you whether the program P will halt on input i. Returns true if it halts and false if it doesn't.

Define a new program G(x) that halts if x does not halt on itself as input and loops forever if x halts on itself as input:

  def G(x):
    if H(x,x):
      while True:
        pass
    else:
      return

If G(G) halts then H(G,G) is true so we end up in the infinite loop, a contradiction. If G(G) does not halt then it returns without looping, also a contradiction.

So our halting oracle H does not exist, so there can not be a function that tells you whether another function will halt on itself as input, so there can not be a function that tells you in the general case whether some other computation will halt, QED.

In principle you can instantaneously set the throttle opening to some position and set the RPM to whatever you want. In time the RPM will rise or fall until the engine is at equilibrium, but throttle position and RPM aren't like mechanically interlocked. Otherwise how could the engine speed up when you go down a hill?

Found it now, thanks, I had to scroll down.

This is like a much cooler version of a thing I made a few years ago for simulating model oscillating engines: https://incoherency.co.uk/oscillating-engine/

How do you tell how much power it is making?

I think the distinction is that you think a life is only worth living if it is a sufficiently enjoyable life, and I think life is worth living intrinsically.

I did skim the Wikipedia article but it didn't seem to engage with anything I was thinking about.

Almost everybody's revealed preference is to stay alive rather than to not be alive, otherwise we'd see a lot more suicide.

making an important choice such as bringing life into the world, without considering the consequences, is already somewhat unethical.

This is a view you can take, but it's not as obviously important as you seem to think.

An alternative view is "making an important choice such as FAILING to bring life into the world, given the opportunity to do so, without considering the consequences, is somewhat unethical."

If the standard was truly that you can't have kids unless you're sure they're going to have a great life, then we would have gone extinct millions of years ago.