Hacker News: yk

New comment by yk in "Sharing AI progress in mathematics"

yk — Wed, 07 Oct 2026 01:15:59 +0000

I'm a transhumanist, I want to build god and kill death. To me this looks like we are moving in the right direction.

So basically I think that the future is getting pretty weird because we are building really powerful tools, though these tools are precisely what allows us to prosper in that future.

New comment by yk in "Nobel Prize in Physics 2026: Francis Halzen"

yk — Tue, 06 Oct 2026 12:03:32 +0000

Pretty cool. One of the big challenges in trying neutrino detection in ice is dealing with funding agencies. The optical properties of ice only get good when you are actually a few kilometers deep in very old ice. That is you can't just build a demonstrator in the nearest glacier with a few leftover photomultipliers, you need to go directly to Antarctica and bore a three kilometer hole into the ice. And projects where the MVP is quite expensive are always very hard to get funded.

New comment by yk in "NASA asked several former SR-71A staffers to help secret restart"

yk — Wed, 30 Sep 2026 11:26:26 +0000

As a SR-71 fan, I fully support the decision.

And as a connoisseurs of project management disasters, I'm going to get popcorn.

New comment by yk in "AMD's random number generator can't generate a 0?"

yk — Tue, 22 Sep 2026 10:36:26 +0000

    return 4 # Determined by fair dice roll.

New comment by yk in "C++26: Trivial infinite loops are no longer undefined behaviour"

yk — Fri, 18 Sep 2026 18:41:02 +0000

I would expect an infinite loop

    while(true) std::this_thread::yield();

to be designed to play nice with the scheduler, while I would assume a infinite loop

    while(true);

to not play nice with the scheduler. Now, I can't really imagine where this matters except for horrible hacky attempts at faking a real time scheduler on windows, but breaking horrible hacky attempts at faking a real time scheduler sounds like the kind of bug you hear about in the evening news.

New comment by yk in "Navier-Stokes – Tristan Buckmaster [pdf]"

yk — Tue, 08 Sep 2026 23:53:33 +0000

The last two points are disputed/sound significantly more reasonable in [0]. So from what I gather, Buckmaster realizes sometime during the call that the biggest result of his career is going to get steamrolled (the blowup of Navier Stokes is a much bigger deal than the blowup of 3D Euler), and on the other hand the openAi guys realize that they are basically talking about internal results with Anthropic and probably have to call corporate right after this call. Between these two stressor the conversation appears to have gone somewhat poorly.

[0] https://x.com/SebastienBubeck/status/2097379411691516310

New comment by yk in "I think the military commissary's freezers were hacked"

yk — Tue, 01 Sep 2026 11:45:24 +0000

Aww, script kiddies first technical document. ... (checks calender) Actually it's 2026, probably someone just typed "Stop me eating ice cream" into claude code instead of the chat interface.

New comment by yk in "Mars Bar from 1991 found – and it's 20g bigger than today's"

yk — Mon, 10 Aug 2026 20:42:09 +0000

Since they downsized apparently in March, it is likely that the sellers on Amazon mentioning 51g very prominently are just selling off old stock.

New comment by yk in "Mars Bar from 1991 found – and it's 20g bigger than today's"

yk — Mon, 10 Aug 2026 18:20:42 +0000

Checking Amazon UK[0], it seems there are 18g, 39.4 g and 51 g Mars bars (and 57 g Mars Protein bars). So it's not clear weather a 60 g Mars bar is that much of a good example of shrinkflation, but 40 g seems to be the wrong comparison.

[0] https://www.amazon.co.uk/s?k=mars+bar

New comment by yk in "Position: LLMs Can't Jump"

yk — Wed, 05 Aug 2026 14:57:08 +0000

The paper is from the 30th of April this year, openAi announced the counter example to the unit distance problem on the 20th of May. That is to say this paper seems to have aged not much but quite poorly.

Workers Drilling in Romania Broke into a Cave Sealed for 5.5M Years

yk — Tue, 04 Aug 2026 16:32:08 +0000

Article URL: https://travelandtannins.com/workers-drilling-in-romania-broke-into-a-cave-sealed-for-5-5-million-years-and-found-an-entire-living-world-that-has-never-seen-the-sun/

Comments URL: https://news.ycombinator.com/item?id=49171223

Points: 138

# Comments: 65

New comment by yk in "There Will Come Soft Rains (1950) [pdf]"

yk — Tue, 04 Aug 2026 15:32:46 +0000

It impressive how prescient the technology feels. I mean in 2020 I would have thought the fully automated kitchen and the voice that asks which poem one wants to hear as fanciful, now it's kinda tech demo level.

New comment by yk in "Prevent cognitive debt by manually retyping LLM-generated code"

yk — Mon, 03 Aug 2026 11:51:41 +0000

Related, a writing advice I stole from Neal Stephenson is to write the first draft by hand. The thing is, there are a lot of small corrections where you kinda should change the text but nah, and if you already committed to copy the entire thing than you are already working at that sentence anyhow.

New comment by yk in ""the very foundation of modern academia has been blown to bits""

yk — Fri, 31 Jul 2026 09:05:53 +0000

The author is a mathematician and I think to a certain extend the tweet reflects the current panic among (some) mathematicians. So ~~second sentence~~ third paragraph the claim that somehow ai could right now write a phd in physics or sociology is something we don't observe (at the moment). What we observe is, that ai can find counter examples to well established conjectures in mathematics quite well, but the thing is the other fields don't have the kind of well established riddles that currently produce the flashy results in mathematics.

New comment by yk in "Glue bonds to nonstick surfaces and wipes clean with ethanol"

yk — Mon, 27 Jul 2026 16:03:40 +0000

On the left hand side we see a lot very unhappy carbon.

New comment by yk in "MIT's New Method Flags AI Models Trained on CASM Without Generating It"

yk — Mon, 13 Jul 2026 23:44:36 +0000

> The approach, detailed in a paper presented at the International Conference on Machine Learning, achieved 100% accuracy in identifying models specialized for CSAM generation.

I know some reasons for 100% accuracy in machine learning, first of all the test set leaking into training data. Or you just accept a silly high false positive rate.

When I was an admin I liked to joke that if you guarantee more than 5 nines, then you are an insurance company and you are planning to pay the penalty instead of actually fulfilling your promise, here the principle is probably the same.

New comment by yk in "When 2+2=5"

yk — Mon, 06 Jul 2026 20:11:11 +0000

Up to a conformal factor.

New comment by yk in "Incident CVE-2026-LGTM"

yk — Fri, 26 Jun 2026 14:47:45 +0000

> Seven LLMs were arranged in series. Six assumed another had read the code; the seventh read it and apologised.

And this is why management assumes that one can just automate software developers.

New comment by yk in "Why Drawing Tablet Brands Won't Collaborate on Linux Floss Drivers"

yk — Mon, 22 Jun 2026 16:04:11 +0000

Pretty sure the reason is, that anybody who could actually do the changes thinks works for me, the alternative is a lot of work for little technical gain.

And besides once you start your tablet for Linux Projekt you have to touch everything, so that is a nice opportunity to finally refactor the wacom_debug_2 mess and pretty soon you're drowning in yak shavings.

New comment by yk in "Frontier AI has broken the open CTF format"

yk — Sat, 16 May 2026 11:50:45 +0000

There's something funny about complaining about cheating in a hacking competition.

Well actually I get it. In cycling motor doping, putting a hidden engine into the bike, seems more offensive than regular doping. I think this is because there is a continuum from eating well to taking supplements to injecting stuff, but having a engine breaks a fundamental idea about cycling. Similar hacking is about cleverly abusing the rules.