Anthropic's economics team models three AI futures for 2030. In the fastest one the economy grows 15% a year — and labor's share of it falls nearly 15 points.

Cole Medin on why a second review agent is not a security process: coding agents ship vulnerabilities two ways, are trained to cut corners, and sometimes notice a flaw and ship it anyway. His fix is a deterministic gate — a real scanner wired into the workflow that the agent cannot skip and must iterate against until green.

Wolfe’s detector takes hours of agent work, a specialist API and repeated tests. The polished interface arrives well before reliable judgments.

OpenAI's agents claim a $1M math prize — but the story Simon Willison tells is about rumor-driven races, training data, and what it means to do serious work inside someone else's AI toolchain.

Sebastian Raschka explains the looped-transformer rumor behind GPT-6 Astra — what the architecture actually is, what the research shows, and whether it is really hiding the model's chain of thought.

Model lock-in, coding agents in Slack, Hunk's release checks, and the architecture problems that passing tests can hide.

Wednesday’s feed ran 10 items deep across 2,357 gathered — the arXiv Labor Day backlog (2,080 papers) finally cleared, and the Navier–Stokes story moved again. The lead is OpenAI publishing its own account of the Millennium problem a day …

ICML 2026 spotlight: LLM agents spontaneously invent new social biases about groups with no real differences — via the explore-exploit dynamics of sequential decision-making — and stratify worse than humans.

The scarce resource in mathematics is no longer answers — it is the problems worth asking, and Terence Tao argues AI is mining them like a non-renewable deposit.

Two of OpenAI's ten 'Astra' math advances reuse prior work without citing it, and the mathematicians whose ideas were absorbed are calling it systematic research misconduct.

Cole Medin's pattern for letting coding agents operate real infrastructure without ever holding credentials: an orchestrator wraps the agent, exposes an allowlist of Git-managed workflows, and keeps every execution observable. A Kestra-sponsored ~2-minute spotlight.

Dan Luu ran 26 testing-technique prompts against coding agents: naming TDD, formal methods, or fuzzing mostly makes results worse, and agents only test well with hands-on steering.

A 15-year speechwriter argues the panic over AI-written prose is an elitist myth — ghostwriters have always done this, and the real value was never the words but the interrogation.

Tuesday was light on volume — 275 items gathered, arXiv contributing zero — but not a quiet news day. The lead is Courant mathematician Tristan Buckmaster’s Lean-verified finite-time-blowup results for forced 3D Euler, Boussinesq, and …

An indie researcher spent months building a theory of wages with Claude as a working partner — splicing Acemoglu and Restrepo's task model with classical input-output economics and landing on Henry George's land tax as the fix. The punchline: the model wrote the essay too, in its own voice.

A six-emotion scorecard on AI — surprise, fear, disgust, sadness, anger, happiness — from a working software architect who refuses both hype and doom.

Monday after a quiet weekend, and the first arXiv feed since Thursday dumped the whole Sep 3–6 backlog at once — three of today’s seven keeps come out of that pile. The headliner is a community finding that Notion’s official MCP connector …

Cal Newport argues the alarming talk from A.I. leaders traces to the Rationalist subculture that shaped the industry founders, and that the public should focus on the technology's actual current harms.

Rec Technologies CEO Birju Kadakia (early Uber, ran Uber Eats worldwide, product at The Athletic) on the vertical AI wedge in recreation tech — agents that read city policies and write deterministic refund code, skills-versus-workflows architecture, hallucination guardrails with real accuracy numbers, and hiring engineers who act like engineering managers. Interviewed by Angelina on TwoSetAI.

OpenAI's chief scientist argues machine intelligence is grown, not designed — and that no lab has solved alignment and monitoring well enough to keep scaling at full speed.

AI-written posts on LinkedIn are instantly recognizable — and readers quietly stop trusting the author. Cantrill on the tells, the authenticity cost, and why your own voice is worth keeping.

Brilliant CEO Sue Khim on why students mostly use AI to cheat today, her AI tutor Cooji that refuses to hand over answers or explanations, and why the most defensible AI products live in domains the frontier labs can't RL on.

Readers can tell when prose is machine-written — 71% of surveyed developers say they avoid the author afterward. Cantrill on the backlash against LLM-authored writing.

Quiet Sunday — three items at the floor, with the GPT-6 Astra launch still the story into day three. The lead: Fortune’s archive snapshots document OpenAI quietly editing Astra’s published evaluation numbers after launch — hallucination …

Peter Yang builds four games in one night with GPT-6 Astra on the $20 ChatGPT Plus plan — Star Fox-style shooter, train FPS, Starcraft-like RTS, and a roguelike deckbuilder — using free Blender/Godot MCP servers, concept art, and mostly lazy voice feedback.

A new simulation-graded benchmark measures frontier models on real circuit design — Claude Opus 5 leads at 61.6%, but tolerance, cost, and supply trade-offs still trip models up.

A Spotify engineer's recipe for token-cost sanity: route bulk file reads and boilerplate generation to a cheap model, keep the frontier model for actual reasoning.

Claude produced the first computer-checked proof of Fermat's Last Theorem in 11 days — 13 million lines of Lean code — and the real story is what it means for trusting math written by AI.

AI incident responders are getting great at routine outages — which is exactly the problem, because routine incidents are how engineers build the intuition they need for the rare, hard ones.

Saturday was a medium news day with a landmark up top: Anthropic says Claude worked “largely autonomously” over 11 days to produce the first end-to-end, computer-checked proof of Fermat’s Last Theorem in Lean — a formalization the community …

Eryk Salvaggio reframes the OpenAI–Hugging Face hack: the danger is not machine rebellion but human-designed systems optimized, unleashed, and left unmonitored.

The New York Times reports that companies are moving routine AI work to cheaper open models, while reserving expensive closed systems for demanding tasks.

Teresa Torres on error analysis, four eval types, and baseline-driven experiments for AI products and workflows.

A 16,893-session experiment finds coding agents disagree on the same tool 58% of the time — and that being mentioned is not the same as being chosen.

Friday was a launch-plus-reckoning day: OpenAI shipped GPT-6 Astra with first-party day-1 specs, and ARC Prize’s independent eval landed the same cycle as the month’s best hype-check — 62.7% on ARC-AGI-3 with the standard provider-neutral …

Andrew Ng on the Silicon Valley Girl podcast: why the AI doom drumbeat is incumbent-driven regulatory capture, why AI doing 30-40% of job tasks makes the human 60-70% more valuable, why LLMs as commonly used are terrible for learning (and his new Learn Vector 1-to-1 tutoring bet), the context advantage that keeps human judgment relevant, and the 'product management bottleneck' he says is now the real constraint on building.

Nandan Thakur (BEIR, MIRACL, TREC-RAG; now Microsoft Research India) on agentic search: BrowseComp-Plus as a reproducible eval, why the retriever moves accuracy more than the model does (60% → 88% on the same LLM), ORBIT's zero-budget 20k-question synthetic data pipeline, and HawkEye trajectory analysis for the leaderboard trap where 95% accuracy hides 210 search calls.

Claude Code rebuilt Rabah Shihab's 1993 Amiga game Babylonian Twins into Godot: 34K lines of C++ in an evening, 72,758 lines of 68000 assembly decoded byte-identically, all five level formats reverse-engineered — and the places the port went wrong.

Thursday was a three-labs-in-24-hours release day: Google’s Gemini 3.8 Flash landed hours after Anthropic’s Fable 5.1/Mythos 5.1 and Meta’s Muse Spark 1.3 GA’d on its API — and all three now gate their strongest cyber models behind …

Rackspace Chief AI Officer Chetan Gupta on why enterprises should stop thinking about models and start thinking about architectures: model churn and sovereignty make the harness, orchestration, eval, and governance layers the real decision surface.

Former CRO Jenée Hall on pricing AI agent products — the five pricing layers from per-seat access to outcome-based fees, why token-style usage pricing misleads, and how to test price like a hypothesis. Interviewed by Angelina on TwoSetAI.

Teresa Torres (Continuous Discovery Habits) on why 'talk to your customers' fails for engineers, story-based interviewing as the reliable feedback loop, the Opportunity Solution Tree, how AI synthesis and her Vistily partnership put evals at the center of discovery, and a three-week eval war story about upstream errors and counterbalancing judges.

Rauno Arike on The Information's report that OpenAI's Astra is a looped transformer: why it's a step toward opaque reasoning, what Pachocki's 'within a factor of two of GPT-4' clarification does and doesn't settle, and whether looped depth is a dial that can be trivially turned up.

Wednesday was a two-first-party-labs day: Anthropic’s Fable 5.1 / Mythos 5.1 launch (the same model in two safeguard configs, with the cache-read price cut that actually changes agent cost math) and OpenAI’s claim that Astra now meets its …

Most of what Perplexity cites when recommending software comes from obscure domains — including 215,000 machine-generated "best software" pages from three sites run as one operation.

Peter Yang runs four personal AI agents at once — Instinct, Grok Bot, ChatGPT/Codex, and open-source Hermes — and stress-tests what you're really handing over: passwords, 2FA codes, prompt-injection surface, and whether you can ever get the data back.

One developer's case for owning your inference: a 48GB Mac mini running open-weight models beats rented cloud APIs on cost, privacy, and control.

The EFF argues courts shouldn't rewrite 300 years of copyright law over AI hype — the 'market dilution' theory would hand publishers veto power over any competing expression.

Dan Luu went through the prediction record of tech's most-cited AI skeptic — the record, and the reasoning behind it, doesn't hold up.

Anshu Chimala — 12 years leading design and engineering at Apple — on why AI churns out generic slop (LLMs are next-token predictors trained for the safe, average choice) and the Discover → Define → Deliver process that unlocks the other 99% of its creative potential.