Hacker News: natsucks

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

natsucks — Mon, 03 Aug 2026 22:31:32 +0000

lol

New comment by natsucks in "LLMs reward expertise"

natsucks — Mon, 03 Aug 2026 21:37:18 +0000

The question i wonder about is, when will an event come along that persuades everyone that human understanding is still required? Or will it never come?

New comment by natsucks in "LLMs reward expertise"

natsucks — Mon, 03 Aug 2026 21:34:55 +0000

And what about problems that cannot be one-shotted but helped along?

New comment by natsucks in "LLMs reward expertise"

natsucks — Mon, 03 Aug 2026 21:30:33 +0000

I am feeling this a lot lately. Getting the most out of agents seems to require being able to ask the right question. And how can you ask the right questions without deep domain expertise?

New comment by natsucks in "Don't be a meat proxy"

natsucks — Mon, 03 Aug 2026 15:57:03 +0000

Claude says this is funny.

New comment by natsucks in "Don't be a meat proxy"

natsucks — Mon, 03 Aug 2026 15:56:14 +0000

Yeah, AI is a general purpose technology, on a par with the introduction of the computer itself. Analogies to tools seems inappropriate.

New comment by natsucks in "Don't be a meat proxy"

natsucks — Mon, 03 Aug 2026 15:51:39 +0000

I have a deep expertise in a domain and when I read that example it was incredibly edifying. I regularly encounter statements that are so concentrated with information that it becomes opaque and I cannot interpret it without reading it over again several times.

New comment by natsucks in "Don't be a meat proxy"

natsucks — Mon, 03 Aug 2026 15:40:14 +0000

oh, interesting idea. edit: https://news.ycombinator.com/item?id=49114639

New comment by natsucks in "Don't be a meat proxy"

natsucks — Mon, 03 Aug 2026 15:36:36 +0000

wow. what a nightmare.

New comment by natsucks in "Qwen3.8-Max: A New Bar for Coding and Cowork"

natsucks — Mon, 03 Aug 2026 15:00:56 +0000

i really like the idea of running local models but i'm always in the position of wanting the best model(s) available and i don't have any severe privacy concerns. as such i have yet to justify ever using local models.

New comment by natsucks in "Superlogical"

natsucks — Thu, 30 Jul 2026 19:59:09 +0000

my main reason: i cannot run agents persistently on a remote host with a desktop app.

New comment by natsucks in "Private healthcare makes industries less innovative. It's time for change"

natsucks — Thu, 23 Jul 2026 15:17:20 +0000

"the thing I miss most is the NHS. It’s been treated like a political football since I left and is apparently a shell of its former self — not because the idea is bad and can’t work but because conservative politicians, some of whom have received funding from private healthcare companies, have deliberately sabotaged it."

really it could not be because government is bad at managing massive programs and provides no incentive for high performance and efficiency....

New comment by natsucks in "Terence Tao's ChatGPT conversation about the Jacobian Conjecture counterexample"

natsucks — Wed, 22 Jul 2026 21:22:20 +0000

More and more the skill of being able to ask the right question seems critical to me, and I don't know how one can do that without deeper and deeper domain expertise.

New comment by natsucks in "GPT-5.6 used a prompt to close a 30-year gap in convex optimization"

natsucks — Sat, 18 Jul 2026 17:57:05 +0000

call me naive, but i really think the pursuit for the new, the better, the interesting, is boundless and humans will never stop. and so for those of us that dream and push and are plain curious will never run out of ways to be useful. and hopefully the fruits of all this tech will be ample enough to support the ones that don't want to or can't participate in that way.

New comment by natsucks in "GPT-5.6 used a prompt to close a 30-year gap in convex optimization"

natsucks — Sat, 18 Jul 2026 17:32:16 +0000

i'd take a yoga class from a bot

New comment by natsucks in "AI Boosts Research Careers but Flattens Scientific Discovery"

natsucks — Sun, 12 Jul 2026 16:34:51 +0000

"Perhaps, says Evans. But he doesn’t think that the problem is baked into the algorithmic design of AI. More than technical integration, he argues, what may matter most is overhauling the reward structures that shape what scientists choose to work on in the first place.

'It’s not about the architecture per se,' Evans says. 'It’s about the incentives.'"

New comment by natsucks in "Please stop the AI confidence theater"

natsucks — Fri, 03 Jul 2026 15:50:28 +0000

Do you think it was the AI itself that gave you the boost here or the discussion and brainstorming?

New comment by natsucks in "Please stop the AI confidence theater"

natsucks — Fri, 03 Jul 2026 15:17:46 +0000

I wish every product everywhere was evaluated based on merit, efficacy, etc alone...but...i bet most of us lean towards products in a grocery store with more appealing labels. The unfortunate reality is that customer psychology matters and your great product can die because of it.

New comment by natsucks in "Midjourney Medical"

natsucks — Thu, 18 Jun 2026 04:04:44 +0000

Not hating, but there's no way resolution gets as good as MRI with ultrasound computed tomography (https://en.wikipedia.org/wiki/Ultrasound_computer_tomography). Doing something like searching for room-temperature semiconductors so that MRI scanners are much cheaper to operate would be a more worthy goal.

New comment by natsucks in "Mouseless – keyboard-driven control of macOS/Linux/Windows"

natsucks — Fri, 05 Jun 2026 19:37:55 +0000

i spent more than year with ergo split keyboards, home row mods, layers, etc. Mouse + speech-to-text + agents made it irrelevant.