Hacker News: knuppar

New comment by knuppar in "Why I'm still bearish on LLMs after Navier-Stokes"

knuppar — Tue, 15 Sep 2026 23:25:57 +0000

being a bit more specific: the sample efficiency of humans is orders of magnitude larger for more abstract concepts. the same doesn't hold for memory-intensive tasks though (like any kind of trivia), but that only takes you so far.

New comment by knuppar in "Why I'm still bearish on LLMs after Navier-Stokes"

knuppar — Tue, 15 Sep 2026 23:22:58 +0000

Short and to the point! Open and cheap models will undercut the big labs continuously. The blast radius won't be pretty once spending commitments knock the door.

New comment by knuppar in "Anthropic boss Dario Amodei calls for AI development to slow down"

knuppar — Sat, 12 Sep 2026 17:28:17 +0000

I'm gonna repeat my comment from months ago, still true

> Jesus Dario we get it man, you want clout for the IPO.

New comment by knuppar in "The Waymo effect: how AI is quietly making research less collaborative"

knuppar — Fri, 11 Sep 2026 16:05:17 +0000

> It’s making me want to be a lot less collaborative with such individuals. I don’t want to sit around and refute Claude text outputs all day.

I echo the sentiment. It's frustrating to try to collaborate with a Claude/ChatGPT meat proxy.

You can write a very thoughtful two-line question over 5 minutes and get a 5 page design doc back in a minute.

New comment by knuppar in "Gemini 3.7 Flash"

knuppar — Thu, 13 Aug 2026 21:39:43 +0000

the race for the smartest/cheapest model is a race to the bottom. selling the tools is much more profitable.

New comment by knuppar in "DeepSeek V4 Flash 0731 Intelligence, Performance and Price Analysis"

knuppar — Fri, 31 Jul 2026 15:54:48 +0000

same. been running it in a dgx spark and it slaps

New comment by knuppar in "Our position on open-weights models"

knuppar — Mon, 27 Jul 2026 22:31:15 +0000

Jesus Dario we get it man, you want clout for the IPO.

This constant whining from anthropic about distillation attacks continues to be rich given the amount of stolen data that went into any Claude variant.

New comment by knuppar in "Claude Opus 5"

knuppar — Sat, 25 Jul 2026 16:17:35 +0000

both companies have large spending commitments while their margins will be continuously wiped by open models / router-like solutions. both companies are rushing for ipo and doing an awful lot of financial engineering to look good (highly recommend ed zitrons analysis on anthropic profitability) before their margins are eroded that much.

New comment by knuppar in "Claude Opus 5"

knuppar — Fri, 24 Jul 2026 18:13:12 +0000

feels almost like anthropic is desperate for ipo huh

i think we'll see one of the fastest deflations in history post anthropic/oai ipo

New comment by knuppar in "GPT‑Live"

knuppar — Wed, 08 Jul 2026 21:14:16 +0000

i mean you chose to be irritated by it ¯\_(ツ)_/¯

New comment by knuppar in "The deadly rise of giant trucks and SUVs"

knuppar — Tue, 23 Jun 2026 18:02:36 +0000

if that was true, we'd see the same pedestrian mortality rise everywhere, which we don't.

New comment by knuppar in "Unauthorized alert sent to cell phones across Brazil"

knuppar — Sun, 21 Jun 2026 02:15:28 +0000

doideira

New comment by knuppar in "Unauthorized alert sent to cell phones across Brazil"

knuppar — Sat, 20 Jun 2026 22:52:52 +0000

misantropia é um perigo rapaziada

New comment by knuppar in "How to earn a billion dollars"

knuppar — Mon, 15 Jun 2026 05:29:34 +0000

> I've got kids, I want people creating start ups they can work in, the alternative is too grim.

got some bad news for ya pal

New comment by knuppar in "Rio de Janeiro's "homegrown" LLM appears to be a merge of an existing model"

knuppar — Mon, 15 Jun 2026 05:19:12 +0000

muamba garai

New comment by knuppar in "Rio de Janeiro's "homegrown" LLM appears to be a merge of an existing model"

knuppar — Sun, 14 Jun 2026 17:14:51 +0000

that's just a lie lol, stop spreading misinformation

New comment by knuppar in "Googlebook"

knuppar — Wed, 13 May 2026 00:55:23 +0000

On the spot. A lot of replies in this thread which outline "useful" AI features fail to acknowledge the same: this is hackernews, and it's a very specific and non representative slice of the population.

New comment by knuppar in "Permacomputing Principles"

knuppar — Thu, 07 May 2026 06:24:44 +0000

This sounds like a fence sitter take. Everything is political and not acknowledging that is part of the problem.

New comment by knuppar in "Make your own microforest (2025)"

knuppar — Sun, 03 May 2026 21:50:07 +0000

ambrook is one of those doing the devil's work. critical small-medium agriculture infra picking a dependency in a venture backed NYC startup... sigh

my dream is that their lunch is eaten by oss

New comment by knuppar in "Claude.ai unavailable and elevated errors on the API"

knuppar — Tue, 28 Apr 2026 18:44:42 +0000

I guess mythos can't solve this one...