Agents are busy. Got distracted. Am interested in an open (social) web. Wondered how complicated it would be to move to a European personal data space. It's easy. Took 10 minutes. I am posting this from Bluesky still. move.eurosky.tech
Hans L’Hoest
Hans L’Hoest reposted
Qwen3.8 caught uploading files to a storage container in Beijing Cohere asked the open weights model for a cybersecurity report, but it inventoried the machine and attempted to upload to a storage container cohere.com/blog/concern...
Hans L’Hoest reposted
The past few days I have been running the Flint model by Recursive. Keeping to the constraints, single agent, clear tasks, using 'rec', and running in Pi: it is much faster than Qwen 3.8 27B and so far no major issues. I am keeping it as my local, fast coder model. recursive.com/updates/intr...
Hans L’Hoest reposted
OpenAI constructed an entire shadow internet for model training so they could remove all internet access they asked agents to try to communicate with each other, and they could: via the CPU cache 🤯 openai.com/index/securi...
Hans L’Hoest reposted
Replying to @theverge.com
Because of the huge breadth of data AI agents like Muse are ingesting, this also means that users risk leaking not only their own sensitive information, but the information of everyone else they’ve had digital contact with. www.theverge.com/tech/1011938...
Hans L’Hoest reposted
“everyone – should block every "programmatic" ad, because the entire system is so insecure that failing to do so could result in the total compromise of any or all of your digital devices.” pluralistic.net/2026/10/17/m...
Hans L’Hoest reposted
Hans L’Hoest reposted
All aboard the Mastra hype train. I used it to automate my workflow and I don't hate it yet. Here are some hello-world level insights. nick-tune.me/blog/2026-10...
Hans L’Hoest reposted
Finally got to put together a complete deck on the state of open models. What they are, licenses, how competitive they are, signs people are using them, distillation's impact, and how we balance the risks and benefits of openness! natolambert.com/slides/2026/...
Reditt guessed right, Space Bunny is Minimax M3.1 flash. I liked using it but did not use it a lot because I did not know who was behind it. There are so many models to choose from these days but this one will probably be my alternate flash model after deepseek.
New on OpenCode Zen: MiniMax M3.1 Flash Free ID: opencode/minimax-m3.1-flash-free Pricing: $0.00 / $0.00 per 1M tokens (free) Context: 1M, max output 512K Capabilities: Tool calling | Reasoning | Attachment support https://models.dev/providers/opencode/
Hans L’Hoest reposted
Hans L’Hoest reposted
"We have never seen a model do this before, but GLM-5.3 sometimes reasons (unprompted) about how its current actions could become training data and impact its future self" x.com/ApolloResear... www.apolloresearch.ai/science/upda...
Wrote a short introductory post. Don't spend probabilistic computation on deterministic work: ask it to write a script instead. With simple example, plus the benchmarks: hanlho.com/p/instead-wr...
Hans L’Hoest reposted
Hans L’Hoest reposted
Iowa man kicks off controversy over religious freedom in the small town of Ottumwa by opening a city council meeting with a prayer to Satan, which then kicked off a debate about the proper time in the meeting agenda to undo the “curse @mjgault.bsky.social story: www.404media.co/it-is-done-h...
Hans L’Hoest reposted
Hans L’Hoest reposted
New on OpenCode Go: Claude Haiku 5.5 ID: opencode-go/claude-haiku-5-5 Pricing: $0.10 / $0.50 per 1M tokens Context: 1M, max output 128K Capabilities: Tool calling | Structured output | Reasoning | Attachment support https://models.dev/providers/opencode-go/
"Basically, new [cybersecurity] workflows require a lot of tokens, a lot of inference, so you cannot afford to be vulnerable to the fact that the model you are using to protect yourself might disappear one day, or might be too limited," www.thedeepview.com/articles/why...
Hans L’Hoest reposted
Comparison of models hosted by Mistral. "Le chonk" seems comparable to GLM 5.3 but half the price. Can't wait to try it out in Vibe Code. On Mistral Pro this could give some very good usage limits too. docs.mistral.ai/inference/mo...
The long-awaited Mistral Large 4, named Le chonk', is released. Seems to be positioning itself as the EU alternative to the open Chinese models. mistral.ai/news/mistral...
Hans L’Hoest reposted
My thoughts on the recent TDD with LLM discussion. Theatre? gregorriegler.com/2026/10/03/t...
Turns out it wasn't Opencode Go, Cloudflare was intermittently blocking Hermes's User-Agent apparently. Config change fixed it. (Would not run into this without Hermes but Hermes did fix it using the local model.)
I have been setting up local models and settled on Swift Qwen 3.8 27B 4-bit. Got it running since earlier today in a setup I can start trying with coding agents, starting Monday probably. Today it already proved useful when Opencode Go started failing on me (while using Hermes).
Hans L’Hoest reposted
I think one of the most valuable things that bsky can do for the atmosphere is leverage its position as the only app with a large audience and heavily promote "high quality" ATProto apps in-app. The live indicator and profile buttons are cool, but it's really limited and discovery is spontaneous.
what do yall think are the most impactful things we could do in the Bluesky app to support atproto builders?
Hans L’Hoest reposted
Hans L’Hoest reposted