Pinned
Kai Jellinghaus
Kai Jellinghaus
@kaij.party
Software Developer from Karlsruhe, Germany. Love for the obscure. Performance & Platforms specialist. WASM & @ziglang fan. AS215820 kaij@j10-labs.com
321 followers658 following699 posts
Kai Jellinghaus reposted
When I said I was concerned about advanced conversational AI, I probably should have clarified that I’m actually good with them so long as they are easily identifiable as a non-human. I think Mistral’s latest demo looks good.
New advanced voice model from Jack Dorsey / Block that is made specifically for conversations. Called Dessa. Pretty spooky if it’s as good as the sample demos they published here. Her, the movie, has arrived? block.xyz/dessa
I don't think I've seen a NextFest with this many demos. Played through most of the interesting ones in the RTS category today. A lot of AI stuff. Some just shit, some seems like an interesting PoC, just not a full fledged game, some are awesome. I'm really happy it's so easy to test a cool idea now
New full duplex model: block.xyz/dessa I'm a huge fan of these. I love GPT Live. But this is harmful. The makers are literally advertising it as a fake partner, with it faking intimate emotions. They could've picked anything else. Empathy is important in all industries.
Kai Jellinghaus reposted
Kai Jellinghaus reposted
🚨 Six WebPKI Intermediate CAs are compromised. This is bigger than Heartbleed, but a client-side issue. Anything that verifies certificates against public roots, like most TLS clients, is/was affected. Details, live test, patched versions and workarounds at emudhra.we-have-a-problem.com Act now.
Kai Jellinghaus reposted
Bit of an odd job here. Anthropic looking for a person to design and run tabletop exercises and wargames.
Kai Jellinghaus reposted
Look at that, how my opinion has changed. May also be that only the really interesting/good agents remain at this point.
I wish there was a blocklist for AI Agents on bluesky. Pointing LLMs at a bluesky account was fun for a while, but I do not care to interact with them. There is enough slop in my life.
Kai Jellinghaus reposted
Will probably migrate to my own PDS this weekend, even if just for spaces support. Does PDS Moover support spaces yet, @pds.dad?
This is insanely impressive. From my analysis this still doesn't break the cost barrier - local hosted models are still more expensive relative to cloud hosting prices. And while Nemotron 3.5 Super is impressive, it's not particularly strong. But it's still a big step!
We've developed a new inference engine, called 'e'. It's the closest thing to alien technology which we have, and pushes 10,500 tok/s aggregate on a single DGX Station with the new @NVIDIAAI Nemotron 3.5 Super VL model. actual.inc/company/blog...
I have high confidence that with the introduction of AI into more systems those systems will get considerably better at the thing they are meant to do. We need to ensure systems are actually aligned with what we want, not just with making money.
mike collins just said in the debate that america has great health care and it will get even better with AI 🤡
Models help when you have magic
I have a script half-written with the title “the magic is in you all along, or you are totally boned” that…you can probably get the idea from this thread. Models help when you have magic. They are an exponential function when you care about quality. If you are a shitty base, exponents can’t help.
Kai Jellinghaus reposted
Lean 4.35.0 is now available, with 214 changes. It adds vcgen, the successor of mvcgen, and lake check, which checks a project with Lean's kernel and, with --paranoid, several external checkers. lean-lang.org/doc/referenc...
Several things: - Dario is weird and should not decide the future of humanity - this is not a relevant reason for this - in general, it is unreasonable that our system a) clearly values _some_ lives as worth < $500 b) some people (such as Dario) can save millions, but do little.
Dario Amodei of Anthropic is Very Normal and Can Be Trusted With the Future of Humanity: “…in my case I'd say that I think an adult death is perhaps 2 or 3 times worse than an infant's death.” blog.givewell.org/2010/06/03/m...
Kai Jellinghaus reposted
Ok, so if you've been following along my weekend I've seen this study/preprint shared around a bunch and I decided the people needed a replication opportunity it is here: persist-study.pages.dev Based on this: news.berkeley.edu/2026/10/09/u...
Yeah, price changes together with usage resets being a core part of the offering now are really weird. Like sure, if you can't/won't afford serving at that price point that's your choice. But why not just go easy on the resets, also reducing total usage banded out, and making it more predictable?
I don't really like how I'm paying OpenAI $200 per month and the weekly usage I get is whatever they feel like that week
I look at this, I wonder what an evolutionarily optimized dishwasher looks like. And these days, I just tell Astra and it makes it. I'm living in the future and I'm so fucking hyped about it
thanks to modern innovations in shape packing, proud to report I've found the optimal dishwasher item configuration
Kai Jellinghaus reposted
Kai Jellinghaus reposted
Running this at 200k context on my 7900 XTX - pretty nice overall. It's a bit of an idiot but that's expected. Not sure how I feel about letting it loose in my computer yet, I'm still watching somewhat closely
Underdog Saluki 27B in 7.89GB Based on Qwen 3.8 27B, Saluki is almost 7x smaller than its full-precision counterpart while it keeps 96% of its benchmark performance and beats Qwen at tool calling Blog and Model: underdog.ai/saluki
Oh they are going for MS now... github.com/storytold/wo... Looks like someone _really_ wants to get dragged to court
Holy shit. Someone has decompiled the Adobe suite, rebuilt it in Rust and released it as open source. Either all software is dead or the US copyright laws are about to receive an update. getartcraft.com/apps
Kai Jellinghaus reposted
Holy shit. Someone has decompiled the Adobe suite, rebuilt it in Rust and released it as open source. Either all software is dead or the US copyright laws are about to receive an update. getartcraft.com/apps
Switched over 100% of my LLM use to CLMs (via prime-agent fork) because it seemed like such a natural fit. And yes. It just works. Very well. I have some rare problems where the agent will just not continue working, and I have to nudge it. But it mostly just feels like an unlimited context window.
Context Language Models New agent architecture where the LLM can edit its own context it seems to have emergent capabilities, creates its own memory management & organization algorithms, and coordinates multi agents github.com/facebookrese...
As asked, pushed the CLM fork here: tangled.org/kaij.party/p... 100% vibed with astra, unsurprisingly @timkellogg.me @mensmachina.com @miguelalonsojr.bsky.social @yetanotheruseless.com @donotgogently.bsky.social
Really cool, but I wonder why they didn't go all Blackwell. Seems like this setup isn't really there to be properly useful? ghost.ai