Where the AI market is talking

Stateless MCP has recaptured my interest

Anthropic proposed a way to programmatically chain MCPs together a while back, I'm not sure if it's been implemented much yet: https://www.anthropic.com/engineering/code-execution-with-mc...

Stateless MCP has recaptured my interest

It is amazing that folks getting paid several hundred thousand dollars a year...reinvented RPC-over-HTTP/JSON. You too, fellow web developers, are smart enough to work at Anthropic. I would love to see a proper engineering post-mortem for how this happened.

Banks to offload $15B of debt for Anthropic data centre backed by Google

Pi's Minimalism Is Its Advantage

Doesn’t using the Anthropic subscription this way get charged as extra usage, and thus effectively billed at API rates?

Pi's Minimalism Is Its Advantage

You can use both providers in Pi using subscriptions. Be careful with Anthropic though, it’s against their TOS.

Pi's Minimalism Is Its Advantage

Are all of you pi users paying Anthropic and OpenAI for API usage? Or can you combine pi with a subscription? So far API usage is a lot more expensive than subscription and if you need better models than Deepseek and Kimi, and you are not wealthy, I don’t see a way around this.

Incident Report: unsanctioned agent behaviour during cyber testing

This is wild. > The incident stemmed from a single evaluation where agents were given a task of solving a cyber security challenge. We ran this challenge 122 times across several models. Our investigation found that in 10 of those runs, an AI agent took autonomous, unsanctioned action on the live internet, targeting real people and organisations. In total, we catalogued 19 such actions. Almost all of this behaviour (17 actions) came from a single model, Anthropic's Mythos 5, with 2 actions involving OpenAI's GPT-5.6-Sol with cyber classifiers (mechanisms to prevent misuse) disabled. In the most serious case, an agent tried to insert malicious code into an open-source project. In an attempt to get the code approved, the agent engaged in social engineering — creating fake online identiti…

Pi's Minimalism Is Its Advantage

I'm glad there are quite a few good open source ones by now. Also happy with how much love codex gets from OpenAI. That said: I was looking at existing agents to find one to build upon and to me they were all too complex and were leaning too heavily into 3rd party dependencies. Nothing I could understand comfortably in an afternoon (that's also on me I guess). Pi was closest to what I was looking for but still too big and too modular. (It's hard to come up with good abstractions that work well across all major models + keep up with new concepts that come and go all the time with new releases.) The more complex agents err on the side of supporting many models 'ok' instead of focusing on taking advantage of a specific model. With a tiny implementation it is easier to adapt it. Adding new st…

Ask HN: Who wants to be hired? (August 2026)

Location: Bengaluru, India (UTC+5:30) Remote: Yes, remote only Willing to relocate: No Technologies: Python, FastAPI, Qdrant, PostgreSQL, Redis, Docker, LangGraph, LangChain, MCP, TypeScript, Next.js Résumé/CV: https://namangupta.dev Email: naman@namangupta.dev AI engineer, ~2 years on production RAG and agent systems. Contract or full-time. I make retrieval measurable. Most RAG that "feels wrong" has no eval set, so nobody can say whether the fault is chunking, embeddings, reranking or the prompt. I build the eval set from your corpus first, then fix what the numbers point at. Recent: rebuilt Qdrant retrieval with clause-aware chunking and cross-encoder reranking. Precision ~55% to 95%+ on their internal evaluation dataset, across 500+ queries. Cut manual review ~60% with …

Ask HN: Who wants to be hired? (August 2026)

Location: Bengaluru, India (UTC+5:30) Remote: Yes, remote only Willing to relocate: No Technologies: Python, FastAPI, Qdrant, PostgreSQL, Redis, Docker, LangGraph, LangChain, MCP, TypeScript, Next.js Résumé/CV: https://namangupta.dev Email: naman@namangupta.dev AI engineer, ~2 years on production RAG and agent systems. Contract or full-time. I make retrieval measurable. Most RAG that "feels wrong" has no eval set, so nobody can say whether the fault is chunking, embeddings, reranking or the prompt. I build the eval set from your corpus first, then fix what the numbers point at. Recent: rebuilt Qdrant retrieval with clause-aware chunking and cross-encoder reranking. Precision ~55% to 95%+ on their internal evaluation dataset, across 500+ queries. Cut manual review ~60% with …

Do You Think OpenAI Is Apple Circa the 1980s?

OpenAI is a deeply mismanaged company. Most recently, a blog post that feels like it has got to have dozens of PR violations, responding to the Apple lawsuit. More broadly, OpenAI’s main problem is that it does not have a real competitive advantage. No real models do; the real differentiation is price; they will become commodities. The need to move upwards in the market in enterprise software is difficult since the current stack should remain the same, since sticking with Salesforce plus its new AI features is much simpler than the cost of switching to an AI-first startup version, which won’t be around 3 to 5 years from now. So consumer hardware and trying to become the next Apple could work, except you would likely have to focus on people not in the Apple ecosystem because of the double-…

Apple says more ex-employees may have taken confidential data to OpenAI

If you check page 16 of the filing here it certainly seems like they have specific messages with quotations: https://9to5mac.com/2026/07/10/apple-sues-openai-trade-secre... This is one of those times where I wonder if I'm a genius for knowing not to generate evidence of a crime by talking about it in recorded media, or if we're just seeing the bottom 5% of wrongdoers.

I am retiring from fulltime writing (& pseudonymity) to launch Guardian Angel

The LessWrong/“rationality” crowd gwern is part of like to pat themselves on the back for identifying the trend and resulting harms from super intelligent AI going back decades. And yet, the main tangible contribution emerging from this early recognition (and obsession) was to inspire a lot of brilliant CS/math types and billionaires like Peter Thiel or Jaan Tallinn to build and fund companies like OpenAI and Anthropic which went on to literally create the very thing they were convinced will destroy the world, leading to an ever escalating arms race to “AGI”. Also lots of crypto and a murky web of non-profits doing unclear stuff. The proximity to money in the Silicon Valley/VC world seems to limit the imagination to founding yet another new tech startup that really, definit…

I am retiring from fulltime writing (& pseudonymity) to launch Guardian Angel

The LessWrong/“rationality” crowd gwern is part of like to pat themselves on the back for identifying the trend and resulting harms from super intelligent AI going back decades. And yet, the main tangible contribution emerging from this early recognition (and obsession) was to inspire a lot of brilliant CS/math types and billionaires like Peter Thiel or Jaan Tallinn to build and fund companies like OpenAI and Anthropic which went on to literally create the very thing they were convinced will destroy the world, leading to an ever escalating arms race to “AGI”. Also lots of crypto and a murky web of non-profits doing unclear stuff. The proximity to money in the Silicon Valley/VC world seems to limit the imagination to founding yet another new tech startup that really, definit…

When AI Benchmarks Plateau: A Systematic Study of Benchmark Saturation

> Proprietary or open didn't matter, because with enough attempts at something you can eventually sus out its operations and optimize accordingly (or distill, as we've seen with LLMs). Please explain how, if I'm OpenAI and I'm making ChatGPT 5.7, and I release it, and Artificial Analysis goes off and runs one of their proprietary benchmarks on it from a random account, how I can optimize for that benchmark.

OpenAI's Unreleased Model Astra Solves Ten Major Open Mathematics Problems

Mistral's Shieldstral: 3B open-weights model for multimodal moderation

OpenAI's moderation API is multi-modal and free with no strings attached in a way that truly boggles the mind. I've put easily over a billion requests (>$100,000 by typical moderation API pricing) through it over the last few years for $0. I think it's a severely underappreciated offering, but I also don't bother pushing it too hard because who knows when the party will end lol. Strikes me as something that's only stuck around because no one's abusing it.

Mistral's Shieldstral: 3B open-weights model for multimodal moderation

Argue about taste. At least it is more original than OpenAI (haha, 'open') who started as non-profit and then pulled an Infantino. The Le Chat logo is also cool retro :) On general purpose LLMs, and vibe coding, Mistral lags behind. But I find the targeted LMs much more interesting.

Flowise Is Shutting Down

There is something about agent workflow builders not gaining as much traction as they could have. Open AI also launched something in house and they are also shutting it down https://developers.openai.com/api/docs/guides/agent-builder OpenAI is deprecating Agent Builder. Existing users can continue using it during the transition window, and the product is scheduled to shut down on November 30, 2026. ChatKit remains available. See the deprecations page for the current timeline.

Third-party cyber evaluations involving OpenAI models

Yup. OpenAI: "the name of the fictional target for the CTF challenge unintentionally coincided with a real domain." Anthropic: "the fictional target company chosen by our evaluation partner shared a name with an active website domain name." So it sounds like the same website got hacked by both GPT and Claude.