Hacker News: agrishin

New comment by agrishin in "Project Glasswing: Securing critical software for the AI era"

agrishin — Tue, 07 Apr 2026 18:31:06 +0000

>>> the US and its allies must maintain a decisive lead in AI technology. Governments have an essential role to play in helping maintain that lead, and in both assessing and mitigating the national security risks associated with AI models. We are ready to work with local, state, and federal representatives to assist in these tasks.

How long would it take to turn a defensive mechanism into an offensive one?

New comment by agrishin in "Show HN: A game where you build a GPU"

agrishin — Sat, 04 Apr 2026 17:35:38 +0000

Great project! I somehow missed whole cpu architecture topic, so gonna catch up on that now

New comment by agrishin in "German men 18-45 need military permit for extended stays abroad"

agrishin — Sat, 04 Apr 2026 16:53:13 +0000

Well, you see, if men stay alive, but women are killed, society collapses eventually as not enough new people are born. It sucks being a man in this scenario, but it is what it is.

New comment by agrishin in "Show HN: Travel Hacking Toolkit – Points search and trip planning with AI"

agrishin — Sat, 04 Apr 2026 09:51:45 +0000

I don't know if US miles gives better deals, but in EU (Flying Blue, KLM) Amsterdam-Munich (1hr flight) business class is 52k miles. Amsterdam - Los Angeles business class goes for 550k miles. For 1 passenger.

New comment by agrishin in "European alternatives to Google, Apple, Dropbox and 120 US apps"

agrishin — Fri, 03 Apr 2026 10:29:19 +0000

Do they really match full functionality and user experience though?

New comment by agrishin in "School uses AI to remove 200 books, including Orwell's 1984 and Twilight"

agrishin — Thu, 26 Mar 2026 23:17:56 +0000

I red 1984 and "Brave new world" roughly at the same time, and for quite some time I thought 1984 to be too unrealistic, and I considered bnw as more likely scenario. I was wrong.

New comment by agrishin in "The Little Book of C"

agrishin — Thu, 26 Mar 2026 23:12:57 +0000

The fact that it's AI generated is simultaneously thrilling and frightening. Especially considering that some AI Agents might be trained on that.

New comment by agrishin in "HyperAgents: Self-referential self-improving agents"

agrishin — Thu, 26 Mar 2026 23:03:34 +0000

I found that running an agent in ralph loop, showing it the agent text and saying "run this, if it fails - identify the reason, and modify the agent instructions to avoid this, acceptance criteria are this and that" worked surprisingly well. Not sure if it qualifies as a self-referential self improving, but it was something.