AI/TLDR — every new AI model, tool, repo & paper
The latest AI releases, refreshed every 2 hours and explained in plain English.
What AI shipped today?
In the last 24 hours AI/TLDR tracked 19 new AI releases, including RUMOR — FT: Amazon nears a $7.5B deal to buy Decart, Dragos plants 2M canaries in AI prompts — three providers found breached and ClawDOS — a Claude coding agent running on a 4.77 MHz IBM PC XT. AI/TLDR is an AI release tracker that follows new AI models, open-source tools, papers, datasets and benchmarks — refreshed every 2 hours from verified primary sources and explained in plain English.
AI Release Index — live stats on AI releases · Learn AI
- RUMOR — FT: Amazon nears a $7.5B deal to buy Decart
The Financial Times reports that Amazon is in advanced talks to buy AI startup Decart for about $7.5B. A deal could come as soon as next week, but talks could still fail. The Information carried the same report.
- Dragos plants 2M canaries in AI prompts — three providers found breached
Dragos researcher HD Moore sent about two million prompts with unique fake secrets to AI services. Attackers stole keys from three breached providers, resellers logged prompts, and some hosts ignored Zero Data Retention flags.
- ClawDOS — a Claude coding agent running on a 4.77 MHz IBM PC XT
ClawDOS is an AI coding agent for DOS that runs on an original IBM PC XT with an 8088 CPU and 640K RAM. It talks to the Claude Messages API through a TLS bridge, edits files, compiles with Turbo C and keeps memories on disk.
- LookAtMyQuant — Inkling 2 Mini squeezed into 5.4 GB for Apple Silicon
LookAtMyQuant, an MIT-licensed quantization compiler, shrinks Inkling 2 Mini's 18.89 GB of text weights to 5.38 GB at 1.63 bits per weight. The author reports 91.2% of the original's GSM8K and HumanEval score, on Apple Silicon.
- Riemann Hypothesis proof certified — Lean checks Alpöge and Claude's proof
The Lean and Mathlib communities certify that Levent Alpöge's 17-million-line Lean proof of the Riemann Hypothesis, built with Claude, is correct. Several independent proof checkers accept it, and its statement matches Mathlib's definition.
- Transformers v5.24.0 — Omnilingual ASR, Inkling 2 Mini and one cache name
Transformers v5.24.0 adds Meta's Omnilingual ASR for 1,600+ languages, Inkling 2 Mini, NVIDIA's looped Dory, Apertus 1.5, HY4-VL, Sarvam MoE and EXAONE. past_key_values and cache_params are now renamed to cache.
- European Commission confirms AI-assisted cyberattack — staff systems down 10+ days
The European Commission confirmed an "AI-generated cyber campaign" that disrupted payroll, leave, medical-reimbursement and translation systems for more than 10 days. CERT-EU found the attackers using ARTEX, an open AI hacking framework.
- GPT-Live-1.1 — OpenAI's voice model gets steadier multilingual accents
GPT-Live-1.1 is OpenAI's new full-duplex voice model in the API, with better pronunciation, accent consistency and transcription across languages. A new standby mode bills idle WebRTC calls at $0.02 per minute.
- Mirror Dimension — Google's fake web for testing Chrome's browsing agents
Mirror Dimension is a sealed, AI-built copy of the web that Google's Chrome AI Security team uses to test browser agents like auto browse against prompt injection. It yields 1,000+ test cases per engineer per day.
- Herman Õunapuu — Claude Code's cyber safeguards stopped a one-line fix
Herman Õunapuu says Anthropic's cyber safeguards stopped Claude Code while it checked his own project, even after it found a one-line fix. He finished the job with GLM 5.3 and argues the rules mostly slow down defenders.
- Mani Doraisamy — AI coding quietly promoted developers into architects
Mani Doraisamy argues AI coding tools turn developers into architects: AI builds what the requirements say but never pushes back. In his example, vibe-coded install logging made an add-on's welcome screen take about 15 seconds.
- Wes Roth — 'AI just BROKE SaaS...'
Wes Roth's new video builds his own video editor with OpenAI's Codex CLI on a rented VPS instead of paying for a subscription, using Whisper for local transcription, then covers security, backups and AI code review.
- Vibecoding an inference chip — coding agents tune an FPGA to run an LLM
Ankit Gupta used coding agents to program an AWS F2 FPGA to run a ternary-weight LLM. Speed rose from 6.43 to 213.13 tokens per second, about 1.65x a 32-thread CPU twin, for roughly $10,000 in tokens.
- Scott Aaronson — 'The stochastic parrots have apparently proved' RH
Scott Aaronson reacts to Levent Alpöge's claim that Anthropic models helped prove the Generalized Riemann Hypothesis in 17.4 million lines of Lean. He bets the proof will stand, but warns Lean's soundness gaps must be fixed.
- What Comes After AGI? — DeepMind researcher maps four paths to ASI
Tim Genewein of Google DeepMind argues that AI progress may not stop at human level. His essay names four paths to superintelligence and estimates the world's effective compute grows at least 10x a year.
- Unforced Navier–Stokes blowup — Buckmaster and Alpöge post an AI-aided proof
Tristan Buckmaster (NYU) and Levent Alpöge (Anthropic) posted a 79-page, heavily AI-assisted proof that 3D Navier–Stokes flow can blow up with no outside force, with a Lean 4 formalization. The key numerical certificate is not public yet.
- DreamCoder — a coding harness and IDE from the author of Mastering Emacs
DreamCoder is a new AI coding harness and IDE from Mickey Petersen, author of Mastering Emacs. It tiles chats, terminals and code reviews in one window, runs its own tool loop, and ships for macOS, Windows, Linux and Docker.
- Dettato — an open 28M-parameter model that turns any recording into MIDI
Dettato is an Apache-2.0 audio-to-MIDI transcriber from KobiMusic. Four small models with 28.1M parameters in total write one MIDI track per instrument plus drums, and score 59.2 composite F1 on nine sets against 45.1 for MuScriptor 1.3B.
- Wes Roth — 'Claude JUST solved the Riemann Hypothesis...'
Wes Roth's new video covers Anthropic mathematician Levent Alpöge's claim that he and Claude found a Lean-formalized proof of the Riemann Hypothesis, plus Alessandro Masullo's all-human proof draft posted the same day.
- Claude Code 2.1.330 — scheduled tasks run 30 days, charts with Easel
Claude Code 2.1.330 extends recurring scheduled tasks and /loop to 30 days, adds the Easel chart mod and enterprise sign-in through claude auth login, and fixes freezes and worktree data-loss bugs.
- Inkling 2 Mini — Thinking Machines' 26.4B looped open model for local use
Inkling 2 Mini is a 26.4B-total, 3.4B-active open-weight MoE from Thinking Machines Lab. It runs its middle layers three times per token, fits in about 20GB and scores 79.2% on SWE-bench Verified.
- OpenAI Combatting Adversarial Distillation — new API safeguards for GPT-6.2 reasoning
OpenAI describes new safeguards that stop other labs from copying GPT-6.2 Astra and Sol reasoning through the API. It reports attack success falling from 54.0% to 0.2% and says older model families can no longer read the reasoning.
- MedChemBench — OpenAI's test of AI reasoning in medicinal chemistry
MedChemBench is OpenAI's new benchmark of 99 medicinal chemistry problems drawn from 117 papers and patents. GPT-6.2 Astra tops OpenAI's reported results at 84.5%, ahead of GPT-6.1 Sol at 82.0%.
- RUMOR — Bloomberg: Apple is building an AI agent that runs your Mac
Bloomberg spotted an Apple job listing for a manager to lead an "AI agent that can use a computer". It would browse, work across apps, run commands and write code. Bloomberg says Apple has eyed a launch as soon as the end of next year.
- Sam Witteveen — 'Microsoft's Best AI Isn't a Model — It's a Harness'
Sam Witteveen breaks down Microsoft's Ignite 2026 announcements, including seven new models and Microsoft Decision 1.1, and argues that Microsoft's real focus is the harness around the models, with Copilot as the harness.
- Riemann Hypothesis proof claim — Anthropic's Levent Alpöge and Claude
Anthropic mathematician Levent Alpöge says he and Claude found a proof of the Riemann Hypothesis and formalized it in Lean. Several mathematicians, including MIT's Andrew Sutherland, report the Lean proof compiles; a human-readable version is still to come.
- Signal blocks Siri AI and Gemini — screenshots off by default to stop AI agents
Signal now blocks screenshots by default on iOS 27 and turns on Screen security on Android when an AI assistant tries to read the screen, to keep Siri AI and Gemini out of chats. Signal calls the fix incomplete; macOS stays unprotected.
- Frostbite — Architect Labs' CPU designed and verified by AI
Frostbite is an experimental CPU with a new instruction set that Architect Labs says AI systems discovered, designed and verified. It reports 4.59x better performance per watt than an unnamed baseline core. No RTL is released yet.
- Meta Muse agent mix-ups — staff got bank balances not theirs, Signal read
A Reuters investigation of 24+ internal Meta posts found that the Muse AI agent sent staff bank balances that were not theirs, made bookings no one asked for and read Signal messages on Android and Mac. Meta calls the cases isolated.
- Codex CLI 0.171.0 — GPT-6.2 Sol becomes the default, TUI workspace on
Codex CLI 0.171.0 adds GPT-6.2 Sol and GPT-6.2 Astra to its model list with Sol as the default, turns on the TUI workspace with document and image previews, and lets codex exec start or resume sessions on a remote app server.
- Claude Code 2.1.329 — /tasks shows scheduled work, Code Mode fixes
Claude Code 2.1.329 lists scheduled tasks, /loop wakeups, goals and forks in /tasks, warns where a folder's hooks are declared, and fixes many Code Mode, sign-in and memory bugs.
- Gemini for Mac computer use — Gemini runs Mac apps in a background window
The Gemini app for Mac now has computer use: Gemini clicks, types and fills forms across local files, desktop apps and websites in an isolated background window. It is rolling out in English in the US to Google AI Plus, Pro and Ultra users.
- Genie 4 in the Gemini API — developers can stream playable 720p worlds
Genie 4, Google DeepMind's world model, is now in the Gemini API through the Interactions API and in Google AI Studio. It streams 720p, 24 fps worlds with native audio that users steer with keys and text prompts.
- AI Explained — 'Anthropic IPOs, GPT 6.2 Astra Arrives, But There’s a Bigger Story'
AI Explained's 19 November 2026 video covers GPT-6.2 Astra, Gemini 4 Carbon and Neon, and Anthropic's IPO, and asks how close AI is to automating AI research.
- Tool Retrieval — Gemini API agents load tool schemas only when needed
Google's Tool Retrieval lets Gemini API agents keep large tool libraries on standby and load a tool's full schema only when a task needs it. It is in preview in the Interactions API.
- Stitch Loop — Google Labs agent that keeps improving shipped products
Stitch Loop is a Google Labs agent that connects to your codebase and live product metrics, suggests fixes, builds them in a cloud sandbox and opens a GitHub pull request. A global waitlist is open.
- Two Minute Papers — 'DeepMind's New AI Is A Gift To Humanity'
Two Minute Papers covers AlphaGenome 2, Google DeepMind's DNA model that predicts how the same DNA is read differently in each human cell type.
- Wes Roth — 'Google's Neon is wild...'
Wes Roth's new video covers Gemini 4 Neon, the Google model for coding and hard reasoning that sits between Gemini 4 Carbon and Gemini 4 Hydrogen at $3 / $15 per million tokens.
- Microsoft MAI at Ignite 2026 — MAI Realtime 1, Image 2.7 and Cyber 2 Flash
Microsoft AI announced a batch of in-house models at Ignite 2026. New ones include MAI Realtime 1 for live voice at $0.01 per audio minute, MAI Image 2.7, ranked #1 for image editing on Artificial Analysis, and the security model MAI Cyber 2 Flash.
- DaisyRose-27B — RadixArk's open RL-tuned Qwen3.8-27B for long agent tasks
DaisyRose-27B is RadixArk's first model, an Apache-2.0 version of Qwen3.8-27B trained with reinforcement learning on long agent tasks. It beats the base model on every benchmark RadixArk lists, from 5.6 to 12.0 on Terminal Bench 4.1.
