AI agents that forget everything between runs, and run actions you never approved? BOTROSTER from @waghweb gives you persistent, named Bots that share one durable computer with a browser, shell and filesystem. Approvals fail closed, a credential broker keeps tokens out of the guest, and every tool call is audited. It's open source and self-hostable, with MCP, Ollama, LM Studio and Linear support.
Weekend hack: HackerTube. I was annoyed with YouTube's recommendations so I had Claude filter down 43M comments over 20 years of HN to 6,454 videos that keep being recommended. *Significantly* better viewing experience, ~2 hours with Opus 5.5:
https://t.co/JIhXN7OSh1
- Excellent recs without any logins (state is kept in local browser session)
- Thumbnail is a representative frame from middle of the video instead of the super duper A/B test hyper-optimized user uploads
- Underlying data from the comment analysis as one zip file, point your Claude at it and ask for your own version
- HackerTube Longs ✊
A bit more on the process:
I use Claude Code (desktop app for Mac), Opus 5.5 on Medium effort as my default. I only talk to Claude and have it do everything, the only "manual work" I helped with in tandem was in the browser (reminiscent of MenuGen): 1) I had to create some kind of API key from Google Cloud UI to query YouTube for video metadata, 2) I booted up a VPS box for serving and 3) I bought the domain name. Once the box was up I gave Claude the IP and told it update, secure, harden, deploy and optimize the code, and it did the whole thing.
Many awesome projects that would have normally taken weeks/months even can now be few-shotted for breakfast. What else are we not thinking of.
I have this great idea for a drug called Regretamine. One dose and all your regrets are gone!
Including, obviously, having taken Regretamine in the first place.
More details about https://t.co/DQoqx0JzSN in Eternal's shareholder letter. It's being positioned as a 'privacy-first'
AI model, with a global launch slated this year.
"Dark uses fine-tuned frontier open-weight models, running inside our own infrastructure.
End-to-end encryption. We cannot read your conversations, even if we are forced to. Context carries across sessions through an encrypted memory management layer you control.
Dark is in the business of keeping you and your thoughts safe. Your thoughts, questions, insights, research, ideas, emotions, everything – stays within your end-to-end encrypted sandbox, so you can talk about anything without any lingering uncertainty in your mind. Because greatness is always born in the dark.
Dark is built for power users – founders, researchers, scientists, enterprises and governments – who want to keep their thoughts private, because privacy is essential for competitiveness. On top of privacy, we have fine-tuned our models to give you the absolute best quality of answers we have seen from any LLM (benchmarks will be revealed soon). Our models are trained to not please you, and tell you what they think is the truth.
Dark will launch globally, hopefully within 2026"
Unfortunately, we are being blocked by certain oligarchs in order to maintain their monopolistic chokehold on the Indian people. You can guess who they are …
This is a crime against the people of India!
Meet Mistral Large 4, aka Le Chonk.
• 1T parameters, natively multimodal. 49B active.
It is the best open weights model from US or Europe on aggregated benchmarks.
• State-of-the-art on critical workloads, including cyber defense, manufacturing and finance and it surpasses closed frontier models on visual grounding.
• Forged in Europe end-to-end and is deployable from Europe via our own Mistral Cloud infrastructure.
• Available to all via API today. Working with cybersecurity partners privately.
Open weights release end of October.