Hacker News: jabedude

New comment by jabedude in "Hospital at centre of child HIV outbreak caught reusing syringes in Pakistan"

jabedude — Fri, 17 Apr 2026 00:31:35 +0000

Are you claiming that Pakistani nurses and doctors are not educated on the dangers of needle reuse?

New comment by jabedude in "The Closing of the Frontier"

jabedude — Sun, 12 Apr 2026 19:30:37 +0000

This happened before with GPT-2 being touted as "too dangerous to release"[0] at the time by OpenAI. I don't think that means every model will be safe to release in the future, but nothing I've read about Mythos seems like it's going to be different this time.

[0]: https://openai.com/index/better-language-models/

New comment by jabedude in "Running Tesla Model 3's computer on my desk using parts from crashed cars"

jabedude — Thu, 26 Mar 2026 03:01:42 +0000

I'm actually somewhat surprised the OS fully boots when it's not connected to the expected vehicle peripherals

New comment by jabedude in "What young workers are doing to AI-proof themselves"

jabedude — Mon, 23 Mar 2026 15:39:20 +0000

Probably a reference to shows like 'Secret Lives of Mormon Wives'

New comment by jabedude in "Meta will shut down VR Horizon Worlds access June 15"

jabedude — Wed, 18 Mar 2026 17:15:38 +0000

At what point does this company un-do their name change to Meta?

New comment by jabedude in "Malus – Clean Room as a Service"

jabedude — Thu, 12 Mar 2026 18:04:06 +0000

This is quite literally the end of open source. projects will find themselves in the position of making their test suites private to avoid being sherlocked like this

New comment by jabedude in "Swiss e-voting pilot can't count 2,048 ballots after decryption failure"

jabedude — Wed, 11 Mar 2026 18:51:20 +0000

Neither is voting free, what's the argument here?

New comment by jabedude in "Statement from Dario Amodei on our discussions with the Department of War"

jabedude — Fri, 27 Feb 2026 17:28:43 +0000

Why should anyone at the Department of War or the general public care what non-citizen employees of these companies think?

New comment by jabedude in "How Taalas “prints” LLM onto a chip?"

jabedude — Sun, 22 Feb 2026 17:45:47 +0000

Just me or does this seems incredibly frightening to anyone else? Imagine printing a misaligned LLM this way and never being able to update the HW to run a different (aligned) model

New comment by jabedude in "Gemini 3 Deep Think"

jabedude — Thu, 12 Feb 2026 18:51:40 +0000

how would we actually objectively measure a model to see if it is AGI if not with benchmarks like arc-AGI?

New comment by jabedude in "Amazon cuts 16k jobs"

jabedude — Thu, 29 Jan 2026 20:28:54 +0000

And Chinese/Korean workers being fired while American workers are being hired by their companies would absolutely be correct to see their jobs being offshored

New comment by jabedude in "Claude Code daily benchmarks for degradation tracking"

jabedude — Thu, 29 Jan 2026 18:08:29 +0000

But that's removing a component that's critical for the test. We as users/benchmark consumers care that the service as provided by Anthropic/OpenAI/Google is consistent over time given the same model/prompt/context

New comment by jabedude in "Amazon cuts 16k jobs"

jabedude — Thu, 29 Jan 2026 02:44:04 +0000

Which is an entirely different dynamic than what the person I responded to was calling for

New comment by jabedude in "Amazon cuts 16k jobs"

jabedude — Thu, 29 Jan 2026 00:58:20 +0000

"Exchanging goods and services for money to a locale" is not a colonial model.

I, a strawberry farmer in Florida, should have no obligation to create an office of locals in every geographic location I sell strawberries in.

New comment by jabedude in "Why does SSH send 100 packets per keystroke?"

jabedude — Thu, 22 Jan 2026 20:17:53 +0000

ssh the protocol doesn't imply any privileges of any kind

New comment by jabedude in "The Startup Graveyard"

jabedude — Tue, 20 Jan 2026 17:59:54 +0000

I don't think this website is implying there was a 'problem' per se. In fact it seems to be geared towards helping resurrect the original premises

New comment by jabedude in "Flux 2 Klein pure C inference"

jabedude — Sun, 18 Jan 2026 23:26:41 +0000

Salvatore, how did you pick up the requisite background knowledge on this subject? IIRC this is your first OSS project in the ML domain, just curious if/how much Claude was helpful with providing you with domain expertise while building this engine

New comment by jabedude in "Rob Pike goes nuclear over GenAI"

jabedude — Fri, 26 Dec 2025 19:12:05 +0000

As it so happens Rob Pike performed absolutely 0 tradeoff analysis

New comment by jabedude in "Rob Pike goes nuclear over GenAI"

jabedude — Fri, 26 Dec 2025 19:10:51 +0000

> And you can't assert that AI is "revolutionary" and "a vague threat" at the same time.

I never asserted that AI is either of those things

New comment by jabedude in "Rob Pike goes nuclear over GenAI"

jabedude — Fri, 26 Dec 2025 14:50:29 +0000

I think it's incredibly obvious how it connects to his "argument" - nothing he complains about is specific to GenAI. So dressing up his hatred of the technology in vague environmental concerns is laughably transparent.

He and everyone who agrees with his post simply don't like generative AI and don't actually care about "recyclable data centers" or the rape of the natural world. Those concerns are just cudgels to be wielded against a vague threatening enemy when convenient, and completely ignored when discussing the technologies they work on and like