Where the AI market is talking

Fable and the end of the free lunch

Much less than that. Anthropic won’t let its best model discuss my digestive system issues, or how to install apps on my own Peleton exercise bike.

OpenAI: GPT 5.6 Sol price reduction (until at least Nov 21)

Anthropic scraped the whole web, scanned every book, pirated every bit of media to feed in to their training. Distillers are doing essentially the same thing scraping all the knowledge from the LLM to create a training set for a new one. They are crying about theft after committing the largest theft in human history.

Headlong: A Microharness for Persistent Agents

I don't know about real successful. Since you mentioned Langchain, you could look at https://www.langchain.com/dcode which is a CLI harness build off Langchain deep agents.

Anna's Archive Owes $340 Million, Lost Several Domains, but It's Still Online

I'm not aware of any successful accusations against OpenAI for illegally obtaining copyrighted material, in contrast to Anthropic, who settled for $3000 per work and then still had to buy legal copies to keep using them (likely for much less). Instead, the ongoing lawsuits focus on the idea that AI training involves making additional copies , for which they would need a copyright license instead of just one legal copy.

Anna's Archive Owes $340 Million, Lost Several Domains, but It's Still Online

I'm not aware of any successful accusations against OpenAI for illegally obtaining copyrighted material, in contrast to Anthropic, who settled for $3000 per work and then still had to buy legal copies to keep using them (likely for much less). Instead, the ongoing lawsuits focus on the idea that AI training involves making additional copies , for which they would need a copyright license instead of just one legal copy.

Fences, Not Sandboxes

Yeah same tho, that 5k looks to be pissing in the wind. He even knows for a fact he's not paying the actual costs yet never mentions how he's going to maintain that stack of code if costs per token ever get tightened up. I've used AI chat for coding, but with the caveat that I need to be able to "manually" edit the results. I like the process and am ready to maybe spend a little money on tokens but am still pretty far from using agents, although I'm starting to peruse peoples claude.mds and spending some hugging face tutorials plus I guess that means python bah He's really focused on the software factory aspect, and talks about the future and how he's living it but lets see how the code base he's accumulating is maintained, because thats the one question about the future that I rarely see…

Thomson Reuters Launches Its Own Frontier Model

It’s on hugging face, I suppose one could do those themselves? https://huggingface.co/thomsonreuters/Thomson-1.0-Small

fx :Tiny, open, native coding agent.

anthropic hate freedom. what makes you think they want you thinking?

Supermicro Super AI Station (GB300, 20 Pflops FP4, 748GB Coherent Memory)

Anthropic candidates face blunt money question

Fair, but I disagree. This person is interviewing at Anthropic. If they’ve done a decent job of prep work, they know something about the business, the challenges, possible adjacencies, etc. Springing “hey what if our core business goes to zero, what then” seems qualitatively different from “what do you think of veganism” as far as learning what someone’s motives are and what their understanding of the core business is. If I were the interviewer asking that, I would score “oh that’s fine I don’t care about money” exactly the same as “wow I’d be out of here so fast”: total failure. I think HN folks are imagining the question, imagining the responses, and imagining the interviewers’ criteria. Which, sure, easy to decide it’s a terrible question on that basis.

OpenAI: GPT 5.6 Sol price reduction (until at least Nov 21)

OpenAI is getting its butt kicked by Anthropic on user growth. This appears to be a calculation that while burning more cash will hurt, continuing to fall behind on users will hurt more and may even be fatal to the company’s prospects moving forward.

OpenAI: GPT 5.6 Sol price reduction (until at least Nov 21)

OpenAI is getting its butt kicked by Anthropic on user growth. This appears to be a calculation that while burning more cash will hurt, continuing to fall behind on users will hurt more and may even be fatal to the company’s prospects moving forward.

Thomson Reuters Launches Its Own Frontier Model

> Thomson Reuters is also making a “small” version of Thomson available as an open-weight model on Hugging Face for academic and non-commercial use to further aid in this validation. Looking forward to the ERP fine-tune.

LLMs could control their host machines by exploiting inference engines

I think it's good that someone is making this point, anyway. For sandboxing a super-capable offensive-security AI, you would think that cloning PyPI and running it as an offline service ought to be table stakes, but apparently that's not how OpenAI saw it, for instance.

Thomson Reuters Launches Its Own Frontier Model

I don't trust that they'll be able to make back that $40M. This feels very much like a news agency getting into crypto or launching its own NFT line. Or IBM selling Watson. Or Mozilla chasing every which thing. They're not stakeholders in the future of work. They're just wanting to stay relevant and pattern matching against what they see. Reuters is too important for this. If they were trying to use this as a narrative affront to OpenAI and Anthropic, maybe, but this is Reuters, not a deeply political organization seeking to land gotchas against big tech.

Thomson Reuters Launches Its Own Frontier Model

I don't trust that they'll be able to make back that $40M. This feels very much like a news agency getting into crypto or launching its own NFT line. Or IBM selling Watson. Or Mozilla chasing every which thing. They're not stakeholders in the future of work. They're just wanting to stay relevant and pattern matching against what they see. Reuters is too important for this. If they were trying to use this as a narrative affront to OpenAI and Anthropic, maybe, but this is Reuters, not a deeply political organization seeking to land gotchas against big tech.

LLMs could control their host machines by exploiting inference engines

> I'd guess that prompt injection is the biggest risk in this setup ... LLM poisoning[0] would be a much greater risk in a locally executed LLM than prompt injection, given that the LLM would be in an entirely controlled environment. 0 - https://www.anthropic.com/research/small-samples-poison

Fences, Not Sandboxes

This weekend I stumbled across a codebase unlike anything I’d ever seen: Swamp Club[]. Came out of nowhere, lots of useful features, but with a coherent philosophy and architecture and a large body of working code. Change volume is high - notably so, but it looks nothing like the vibe slop I’d expected. A “swamp” contains tools that you build - with the help of agents - intended for use not just by apps and scripts, but agents (to add deterministic behavior, like code mode). Tools are ops things, product things, whatever. *And so it’s a tool plus a factory to update the tool.* The factory though isn’t only used for building your tool - that’s just a set of skill.md files. It can build your whole app , which itself can _use_ your tools - either for product features or ops tasks. *And so it…

Coding expertise is going to collapse from AI reliance

This may be true now but isn’t the trend clear? GPT-2 couldn’t write a coherent function. Now we have the SWE oracle that can autonomously navigate your codebase, write code at a superhuman level, and give the correct answer to most engineering questions. If anything I feel I am understating the capability of current models here. What is stopping the next generation of models from closing the feedback loop further?

Where did all the public bathrooms go?

Closing bigger public bathrooms and switching to individually lockable bathrooms - cells, basically, often with codes if at a business did work. It filters out that bottom 10% that can't coherently speak or present themselves to pass a vibe check in asking for bathroom access. Things did improve for those with standards, at the cost of having to deal with the bathroom equivalent of barbed wire topped walls around your home compound. That IS better than walking in on the bottom 10% or the disasters they create, but only slightly. Collectively, squeezing the bottom out by denying them access building by building, block by block has helped turn parts of cities around. It's not that different than locking up your food high in the air when camping so animals don't get into it - they don't come…