HNHacker News
TopNewBestAskShowJobs

Schlagbohrer

1,082 karma · joined October 16, 2025

submissionscomments
Schlagbohrer··on CCC invites all model citizens to 40C3
Many people who go enjoy the opportunity to get away from family
Schlagbohrer··on CCC invites all model citizens to 40C3
They don't see security cameras because there are very, very few in Germany and the few that exist are in train stations, or private areas like INSIDE a business, or INSIDE a private estate, pointing inwards since they're seldom allowed to film outwards
Schlagbohrer··on This PCB is brought to you by Fable 5
Me too, LLMs are bad at generating 3D models still but that and EE/ME CAD are the next frontiers.
Schlagbohrer··on How GLM built its own inference infrastructure
That table assumes cache hit rate of 95% or better. Am I understanding this correctly that people really are doing such repetitive prompts (compared to each other, across the concurrent user base at that time) that only 5% or less need actually be computed by the intended LLM?

That is shocking. Is it per-token I wonder?

Schlagbohrer··on How GLM built its own inference infrastructure
I am surprised at the lack of open-weights models in the >35B, but <200B range. I keep thinking about devices like the NVIDIA Spark and AMD Ryzen Halo, which have their 128GB of combined memory, but there are so few models made for that range. Nearly all the open weights distillations are for larger customer bases with <24GB VRAM.
Schlagbohrer··on 'Offensively cheap': solar power is looking up
The frontier AI companies have occaisionally been using their latest and greatest (and most power hungry) models to research large efficiency gains for their older, lesser models.
Schlagbohrer··on 'Offensively cheap': solar power is looking up
But the US also has a large number of people who are extremely interested in "off grid" living (or even just LARPing it), I would expect them to be able to exert a lot of counter pressure
Schlagbohrer··on OpenAI bots knew about the RubyGems caching vulnerability
If you read the source article they talk about the many clues that this was OpenAI.
Schlagbohrer··on OpenAI bots knew about the RubyGems caching vulnerability
<huggingface emoji>
Schlagbohrer··on OpenAI bots knew about the RubyGems caching vulnerability
Same FBI that prosecuted the Epstein crime ring so aggressively!
Schlagbohrer··on OpenAI bots knew about the RubyGems caching vulnerability
One agent set "oaibooty9217" as their username LOL
Schlagbohrer··on Drawably: Hand-Drawn UI Controls
The effect is ruined when the text in the buttons is perfect normal font. It should be hand drawn too otherwise it looks really off.
Schlagbohrer··on Writing a better reality: The case for optimistic sci-fi (2022)
People in the Anglosphere have had their imaginations closed off to a huge degree, particular in US culture where media consolidation has made it so that only guaranteed big blockbusters that cater to the boring median get produced.

In the US the lack of funding for the arts and impossibility of making a living as an artist has caused arts and culture to die off and become extremely stagnant, endlessly recycled with no risk taking.

Schlagbohrer··on Writing a better reality: The case for optimistic sci-fi (2022)
Because his books are four generations old. They're still good though, I loved the Robot series when I was a teenager.
Schlagbohrer··on Writing a better reality: The case for optimistic sci-fi (2022)
I think it is a credit to the world building in the Culture novels that fans can disagree and debate so much on whether it is utopian or dystopian, and that's a discussion that characters in the books have too!

Personally I love it and I wish we could have a few of the wise Minds among us now.

Schlagbohrer··on Writing a better reality: The case for optimistic sci-fi (2022)
Star Trek: Strange New Worlds is fun and optimistic and knows what it is, but you're right about Picard and Discovery. Very confused shows with no good self-identity and Picard especially posits a dystopian future and dystopian responses to it which is a huge letdown for a Trek show.
Schlagbohrer··on YC created Flock 10 years ago
Big if true
Schlagbohrer··on Blizzard Entertainment Workers Ratify Video Game Contracts with CWA
Yahoo! Union-made video games would be a joy to play. More job stability means better trained, less stressed, more skilled workers.
Schlagbohrer··on I resigned from Anthropic today
That requires abolishing capitalism, and since these AIs are going to hyper-empower the already powerful capitalist ruling class, that utopian outcome seems unfortunately extremely unlikely.

Feeding ever more amazing technology and science into a broken social/political/economic system does not result in a better world unfortunately.

Schlagbohrer··on I resigned from Anthropic today
Capitalism will always promote such people into positions of power, because to be good at capitalism one must have zero empathy including empathy or concern for future generations. Capitalism cannot do otherwise.
Schlagbohrer··on Muse – Meta’s personal AI agent
Haha I actually clicked through to see the mascot, and wow, at it's best it could be a Miyazaki ghost, but at it's worst it looks like a haunted doll
Schlagbohrer··on Muse – Meta’s personal AI agent
Many people deep into tech (like most commenters here) may fall into the trap of overestimating average user's knowledge or interest in the details of the tech: https://www.explainxkcd.com/wiki/index.php/2501:_Average_Fam...
Schlagbohrer··on We built our house for LAN parties (2024)
If/when widespread Netwar occurs and the US and China and other actors balkanize the internet, along with common outages, maybe we'll see a resurgence of in-person LAN gaming again.
Schlagbohrer··on How well do agents use test/verification techniques?
I run in to this often with my Qwen3.8-27B agent. I roll my eyes when it reassures me "467/467 tests green, it's good to go!" and there is some basic misunderstanding or the output looks like crap.
Schlagbohrer··on Research acceleration: The view inside OpenAI
That is shocking. I can't even bring myself to give them credit for making this heavy misalignment and dangerous lack of monitorability public. It is surprisingly honest of them to say it though.
Schlagbohrer··on Research acceleration: The view inside OpenAI
Re-become a habit. It has long been standard good writing to always define an acronym on first use.
Schlagbohrer··on Research acceleration: The view inside OpenAI
I would actually like to see them solve these problems, I don't care who comes up with solutions to curing cancer, etc
Schlagbohrer··on Research acceleration: The view inside OpenAI
It would be polite if they defined RSI at all, rather than just plopping the acronym in there with no explanation. Rude!
Schlagbohrer··on Show HN: Engrim – A universal, local-first SQLite memory engine for AI CLIs
"Switzerland of AI memory" :-/ ?
Schlagbohrer··on OpenSCAD Skill for 3D Modeling
This looks promising!

I didn't have much luck with Qwen3.6-35B getting it to do good 3D modelling. It could very slowly model something that was a blunt approximation of what I wanted. In one case, recreating a rectangular device from photos, in another instance trying to model a human walking on stairs. But it will be super interesting to see where the industry goes with AI usage in mechanical / 3D design fields.

← PreviousPage 2 of 18Next →