Hacker News: snarkconjecture

New comment by snarkconjecture in "Be skeptical of OpenAI's rogue hacker agent story"

snarkconjecture — Fri, 24 Jul 2026 23:16:27 +0000

I think the most important question here is whether Huggingface was "in on it" – whether they knew in advance that this was going to happen and agreed to it. Second most important is whether OAI caused this to happen on purpose, with HF as an unwilling target.

Three options:

(1) Both companies are lying through their teeth and the entire thing is bullshit

(2) OpenAI hacked another company's servers and deliberately gambled on felony charges for the sake of PR

(3) OpenAI is telling the truth and a rogue AI agent hacked another company, opening them up to legal liability accidentally

All three of these are a lot more interesting than the "skeptical" story that this is just OAI doing business-as-usual PR hype, nothing to see here, move along.

(FWIW, I strongly doubt HF was in on it.)

New comment by snarkconjecture in "Claude Fable produced a counterexample to the Jacobian Conjecture"

snarkconjecture — Mon, 20 Jul 2026 21:11:19 +0000

It's absolutely not a trivial combination of prior work, or mathematicians would've found it already. This was not an obscure conjecture.

New comment by snarkconjecture in "Claude Fable produced a counterexample to the Jacobian Conjecture"

snarkconjecture — Mon, 20 Jul 2026 21:08:52 +0000

Checking the counterexample is a Calc II problem. Finding it was beyond the efforts of the best algebraic geometers for almost a century.

New comment by snarkconjecture in "Claude Fable produced a counterexample to the Jacobian Conjecture"

snarkconjecture — Mon, 20 Jul 2026 21:06:41 +0000

Current LLMs can absolutely multiply floats without a tool call. In fact, that's a much more rote symbol-manipulation task than doing original math research.

New comment by snarkconjecture in "A stray "j" ruined my evening"

snarkconjecture — Sun, 28 Jun 2026 14:51:54 +0000

No, it's ^J because `J` is 0x4A in ASCII and `\n` is 0x0A, just as `I` is 0x49 and `\t` (tab, ^I) is 0x09.

In the stone age, pressing CTRL flipped that bit, so ^J is literally "ctrl-J".

New comment by snarkconjecture in "90% of the T Distribution"

snarkconjecture — Sun, 31 May 2026 01:37:30 +0000

Unless I'm mistaken, this uses "standard deviation" to refer to standard error throughout. They differ by a factor of sqrt(num_samples).

This is actually much more commonly useful than the t distribution, in my experience. You can squint at a histogram (or some summary stats), eyeball the stdev, approximate the stderr in your head, and get a pretty good sense of confidence.

I most often find myself doing this for the Bernoulli distribution, where it's also handy to know that the stdev is sqrt(p(1-p)), or "about 1/2 if p is middling, or sqrt(p) when it's small" (and you can flip the polarity to handle p→1).

New comment by snarkconjecture in "Shakespeare’s World – I thought this would be simple but"

snarkconjecture — Sat, 30 May 2026 21:42:20 +0000

The map contains a bunch of references to America, the West Indies, Guiana, and Mexico. (Often with a connotation of "faraway exotic place" or "exciting new international development".)

He may not have written about the British colonies but the New World was clearly at least somewhat present in his mind and his audience's minds.

New comment by snarkconjecture in "Belgium stops decommissioning nuclear power plants"

snarkconjecture — Thu, 30 Apr 2026 18:43:57 +0000

> These reactors can be made safer, but they all still have a foundational design flaw which means the ultimate goal should be replacing rather than continually spending money reinforcing.

This was about the Fukushima reactors that were completely destroyed? In response to a discussion of Belgian reactors that are completely different?

New comment by snarkconjecture in "I cancelled Claude: Token issues, declining quality, and poor support"

snarkconjecture — Fri, 24 Apr 2026 19:53:35 +0000

> the underlying working principles are the same as GPT-2

I don't think anyone was claiming otherwise. Sonnet is still better at writing code than GPT-2, and worse than Opus. Workflows that work with Opus won't always work with Sonnet, just as you can't use GPT-2 in place of Sonnet to do code autocomplete.

New comment by snarkconjecture in "Framework Laptop 13 Pro"

snarkconjecture — Tue, 21 Apr 2026 20:05:24 +0000

I am much more likely to fault them for omitting important information specifically to hide a weak point of the product rather than out of laziness.

New comment by snarkconjecture in "Quine Game"

snarkconjecture — Thu, 02 Apr 2026 07:21:16 +0000

Write a Python quine with three or four hands tied behind your back

Quine Game

snarkconjecture — Thu, 02 Apr 2026 07:21:16 +0000

Article URL: https://adam.scherl.is/static/quine-game/

Comments URL: https://news.ycombinator.com/item?id=47611070

Points: 1

# Comments: 1

New comment by snarkconjecture in "Too Much Color"

snarkconjecture — Fri, 20 Mar 2026 10:12:03 +0000

Computer screens have three-dimensional color spaces. Tetrachromacy doesn't change that.

New comment by snarkconjecture in "Show HN: What's my JND? – a colour guessing game"

snarkconjecture — Wed, 11 Mar 2026 05:28:09 +0000

Tetrachromacy wouldn't affect a test taken through a phone screen.

New comment by snarkconjecture in "Billion-Parameter Theories"

snarkconjecture — Tue, 10 Mar 2026 23:55:56 +0000

Deep neural networks can generalize well even when they're far into the overparametrized regime where classical statistical learning theory predicts overfitting. This is usually called "double descent" and there are many papers on it.

New comment by snarkconjecture in "Why do people keep writing about the imaginary compound Cr2Gr2Te6?"

snarkconjecture — Wed, 27 Aug 2025 03:05:16 +0000

It's more like saying pi is approximately "3..14". Easily corrected syntax errors aren't as bad as semantic errors.

Hex Runes

snarkconjecture — Sat, 01 Mar 2025 03:16:51 +0000

Article URL: https://adamscherlis.github.io/blog/hex-runes/

Comments URL: https://news.ycombinator.com/item?id=43215384

Points: 2

# Comments: 0

New comment by snarkconjecture in "Hot take: GPT 4.5 is a nothing burger"

snarkconjecture — Fri, 28 Feb 2025 23:49:11 +0000

Versions numbers for LLMs don't mean anything consistent. They don't even publicly announce at this point which models are built from new base models and which aren't. I'm pretty sure Claude 3.5 was a new set of base models since Claude 3.

What do mean by "it's a 1.0" and "3rd iteration"? I'm having trouble parsing those in context.

New comment by snarkconjecture in "Iterated Log Coding"

snarkconjecture — Wed, 26 Feb 2025 21:06:53 +0000

Not really. Dirac's trick works entirely at a depth of two logs, using sqrt like unary to increment the number. It requires O(n) symbols to represent the number n, i.e. O(2^n) symbols to represent n bits of precision. This thing has arbitrary nesting depth of logs (or exps), and can represent a number to n bits of precision in O(n) symbols.

Iterated Log Coding

snarkconjecture — Wed, 26 Feb 2025 07:43:21 +0000

Article URL: https://adamscherlis.github.io/blog/iterlog-coding/

Comments URL: https://news.ycombinator.com/item?id=43181610

Points: 109

# Comments: 36