Hacker News: fleischhauf

New comment by fleischhauf in "Using LLM for confidence scoring is a bad idea"

fleischhauf — Thu, 22 Oct 2026 07:47:54 +0000

that's the annoying part about LLM providers, they don't tell you anything if the model changes behind the API. I hope open models catch up

New comment by fleischhauf in "Coding is not solved"

fleischhauf — Tue, 29 Sep 2026 12:26:31 +0000

can't be sure the development continues. As much as opus got better at programming the less I understand the output. it keeps inventing terms and skipping steps for the reader to understand what is written properly. I also feel like progress in programming is becoming a lot slower now.

New comment by fleischhauf in "The Citizen Developer"

fleischhauf — Thu, 20 Aug 2026 22:38:10 +0000

is it going to be like YouTube, where everyone was a video artist suddenly?

New comment by fleischhauf in "We're Getting the Wrong Message from Mythos"

fleischhauf — Fri, 10 Apr 2026 14:24:15 +0000

didn't they also claim this about gpt-2? for sure there is a lot of PR involved as well. Models can also be both, really good at cyber security and bad at writing emails.

New comment by fleischhauf in "Claude Code is locking people out for hours"

fleischhauf — Tue, 07 Apr 2026 17:14:23 +0000

I'm quite impressed on how far they got while the claude code code looks like it does.

New comment by fleischhauf in "The "are you sure?" Problem: Why AI keeps changing its mind"

fleischhauf — Mon, 16 Mar 2026 20:37:48 +0000

I wonder what happens if you prompt it to be a tool and not an assistant and that it does not need to be helpful just do as instructed or something like this

New comment by fleischhauf in "I Used Claude to File My Taxes for Free"

fleischhauf — Tue, 03 Mar 2026 19:20:02 +0000

I wonder how well this does on a German tax declaration, might be a good alternative to touring test judging by complexity

New comment by fleischhauf in "Claude's Cycles [pdf]"

fleischhauf — Tue, 03 Mar 2026 17:10:30 +0000

this. could already be useful to narrow down the search space

New comment by fleischhauf in "European Commission issues call for evidence on open source"

fleischhauf — Fri, 09 Jan 2026 13:31:08 +0000

something like a European GitHub you mean? (didn't read the article)

New comment by fleischhauf in "GPT-5.2"

fleischhauf — Fri, 12 Dec 2025 12:02:37 +0000

I'm always impressed how fast people get used to new things. couple of years ago something like chatgpt was completely impossible, and now people complain it something's does mit do what you told it to and sometimes lies. (not saying your points are not valid or you should not raise them) Some of the points are just not fixable at this point due to tech limitations. A language model currently simply has no way to give an estimate of its confidence. Also there is no way to completely do away with hallucinations (lies). there need to be some more fundamental improvements for this to work reliably.

New comment by fleischhauf in "How elites could shape mass preferences as AI reduces persuasion costs"

fleischhauf — Thu, 04 Dec 2025 21:02:41 +0000

the way people treat Llms these days is that they assign a lot more trust into their output than to random Internet sotes

New comment by fleischhauf in "Germany to ban Huawei from future 6G network"

fleischhauf — Fri, 14 Nov 2025 19:30:05 +0000

laughs in unhinged head of state that imposes arbitrary tarrifs on half of the world and mistakes dementia test for IQ test, says it's difficult

New comment by fleischhauf in "AI for Senior Software Engineers"

fleischhauf — Sun, 09 Nov 2025 10:09:21 +0000

why would it not be fine if the content is fine but it's fully AI generated? Just curious on why that would not be on with you

New comment by fleischhauf in "I'm drowning in AI features I never asked for and I hate it"

fleischhauf — Sun, 26 Oct 2025 04:38:51 +0000

that's already happening for Juniors, combination of ai and less funding

New comment by fleischhauf in "Mesh2Motion – Open-source web application to animate 3D models"

fleischhauf — Fri, 24 Oct 2025 12:21:49 +0000

why? aren't you desperately longing for trying to put into a prompt how a model should be animated? /s

New comment by fleischhauf in "What Happened in 2007?"

fleischhauf — Sun, 19 Oct 2025 14:46:14 +0000

keyword is mostly here. it opens up the opportunities where there was really nothing before.

New comment by fleischhauf in "ChatGPT Pulse"

fleischhauf — Thu, 25 Sep 2025 19:31:09 +0000

how strong are you ?

New comment by fleischhauf in "If you are good at code review, you will be good at using AI agents"

fleischhauf — Sat, 20 Sep 2025 10:37:56 +0000

can I ask what language you are using AI for, there is also a difference in performance for AI in different languages

New comment by fleischhauf in "Chat Control Must Be Stopped"

fleischhauf — Mon, 08 Sep 2025 23:17:04 +0000

as far as I understand they want the software on your device, at one point you need to decrypt if you want to read the message content

New comment by fleischhauf in "We’re Not So Special: A new book challenges human exceptionalism"

fleischhauf — Wed, 20 Aug 2025 08:54:18 +0000

from an intelligence perspective the human brain is ultra efficient