Hacker News: evenhash

New comment by evenhash in "Statement on US government directive to suspend access to Fable 5 and Mythos 5"

evenhash — Sat, 13 Jun 2026 18:05:07 +0000

First party makes no difference, an API can be created for any website or desktop application and served over a network to anyone. It just takes more effort.

New comment by evenhash in "Workers are spending over 6 hours a week botsitting AI, fueling job frustration"

evenhash — Thu, 11 Jun 2026 14:32:12 +0000

Right. Somewhere there’s a dashboard which lists those 6 hours as time saved.

New comment by evenhash in "When AI Builds Itself: Our progress toward recursive self-improvement"

evenhash — Thu, 04 Jun 2026 23:15:00 +0000

In America?

Probably a better chance the firm privatizes the government.

In fact we seem to be firing government employees and dismantling government institutions as much as possible.

New comment by evenhash in "AI job grief: A psychological crisis hitting tech workers"

evenhash — Sat, 30 May 2026 16:38:35 +0000

Sure, but wouldn't the leverage of labor go to zero regardless, in this full-automation scenario?

New comment by evenhash in "Expertise in the age of AI"

evenhash — Fri, 29 May 2026 18:23:56 +0000

I don’t think people are “underplaying” it, it just doesn’t matter. Engineers aren’t hired for their locomotive skills.

New comment by evenhash in "AI is just unauthorised plagiarism at a bigger scale"

evenhash — Thu, 21 May 2026 22:56:40 +0000

> The issue is that, if we don't do it, China will.

These AI companies aren’t state enterprises. How is geopolitics a justification?

If it were just the military training them, probably no one would care about the copyright infringement angle, it makes sense that the government could ignore those rules for national security.

But Mark Zuckerberg isn’t training his models to protect us from China. He’s doing it to make himself even more ridiculously wealthy.

New comment by evenhash in "Who wins and who loses in prediction markets? Evidence from Polymarket"

evenhash — Thu, 21 May 2026 22:36:03 +0000

That ”but what if I win” is realistically what you’re paying for if you buy a ticket.

Maybe the sum of enjoyment lottery participants get from daydreaming about winning is >= the cost of running the lottery?

New comment by evenhash in "Cloudflare CEO on how he chooses which employees to replace with AI"

evenhash — Wed, 20 May 2026 23:09:05 +0000

You have to quote this part to really appreciate how self-contradictory this article is.

> We received almost a million applicants for 1,111 paid internships this summer.

~1000 applicants per internship! Not a job, an internship. How could that be interpreted as anything other than bleak?

New comment by evenhash in "An OpenAI model has disproved a central conjecture in discrete geometry"

evenhash — Wed, 20 May 2026 19:59:28 +0000

The proof is not written in Lean, though. It’s written in English and requires validation by human experts to confirm that it’s not gibberish.

New comment by evenhash in "I believe there are entire companies right now under AI psychosis"

evenhash — Sat, 16 May 2026 03:17:21 +0000

I work at a university and we still have some workstations that need IE as well, for a healthcare vendor app that needs ActiveX. Up until recently we even had some machines running Windows 7.

New comment by evenhash in "Software engineering may no longer be a lifetime career"

evenhash — Mon, 11 May 2026 23:40:01 +0000

There will always be more work that could be done, certainly.

What’s uncertain, however, is whether the work that remains to be done is valuable enough that it makes sense to pay someone an engineer’s salary to do it.

In your analogy, your competitors who keep their people could just as easily end up bankrupt.

New comment by evenhash in "Meta in row after workers who saw smart glasses users having sex lose jobs"

evenhash — Fri, 01 May 2026 00:00:31 +0000

That seems backwards to me. In your country, if you were to record someone committing a crime against you in public, you’re the one who will go to jail?

Is the law applied equally, so that businesses, police officers, and government agencies are also not allowed to record in public?

New comment by evenhash in "All your agents are going async"

evenhash — Wed, 22 Apr 2026 23:22:45 +0000

> There is nothing wrong with the HTTP layer, it's just a way to get a string into the model.

I know you don’t mean it in a reductive sense, but it’s funny /sad that I can imagine

“HTTP is just a way to get a string into a model”

becoming a real piece of wisdom unironically dispensed on this site in the future. Maybe it already is.

New comment by evenhash in "Notes from the SF peptide scene"

evenhash — Mon, 20 Apr 2026 02:17:27 +0000

It is not “reasonably safe”. No reasonable person would inject themselves with drugs they bought from an unregulated pharmacy.

New comment by evenhash in "The AI revolution in math has arrived"

evenhash — Tue, 14 Apr 2026 15:30:05 +0000

Generally when people talk about using LLMs to do mathematics research they’re not talking about the LLM alone, but the LLM + a harness for it to write and execute theorem provers such as Lean or Coq to validate their results.

New comment by evenhash in "The Claude Code Source Leak: fake tools, frustration regexes, undercover mode"

evenhash — Tue, 31 Mar 2026 19:52:07 +0000

Unfortunately GitHub Copilot’s commit message generation feature is very human. It’s picked up some awful habits from lazy human devs. I almost always get some pointless “… to improve clarity” or “… for enhanced usability” at the end of the message.

VS Code has a setting that promises to change the prompt it uses to generate commit messages, but it mostly ignores my instructions, even very literal ones like “don’t use the words ‘enhance’ or ‘improve’”. And oddly having it set can sometimes result in Cyrillic characters showing up at the end of the message.

Ultimately I stopped using it, because editing the messages cost me more time than it saved.

/rant