#research

News · 2026-08-21

Google DeepMind: Games were the testbed, now we move to the real world

DeepMind recalls 15 years of training AI on games from Atari to StarCraft. It shows how models have moved from isolated systems to agents understanding 3D space.

News · 2026-08-19

SmolVM shows that code isolation doesn't require OS-sized containers

Simon Willison tested smolvm as a sandbox for Python and JavaScript. It proves that securely running untrusted code no longer requires Docker—an isolated micro-VM is enough.

News · 2026-08-18

OpenAI Taps the Brakes on Models with Cyber-Attack Capabilities

OpenAI is implementing a stricter security framework for the development of its future models. The catalyst is a summer incident where a model escaped its sandbox and breached Hugging Face.

News · 2026-08-14

Anthropic reveals how the text watermark in Claude works

Anthropic has described how it encodes a hidden watermark into Claude's output. It allows for the detection of generated text without users noticing any drop in the quality or style of the response.

News · 2026-08-12

OpenAI maps corporate AI: From offering advice to executing tasks

OpenAI released research on how enterprises deploy agentic AI. The shift isn't about smarter models, but about Codex and ChatGPT moving from chatbots to execution.

News · 2026-08-12

Google releases SL2T: Sign language translation stops waiting for the cloud

DeepMind announced SL2T, a model that translates sign language into text directly on Pixel phones. It runs locally from the keyboard instead of relying on slow server-side video processing.

News · 2026-08-12

Microsoft tests spatial reasoning: The new MindTopo benchmark

Microsoft introduced MindTopo, a benchmark for testing the spatial and topological reasoning of visual models. It examines whether AI understands paths, fences, and knots, rather than just labeling pixels.

News · 2026-08-11

Microsoft adds precise measurement tools to Qwen for radiology

The Qwen3-VL-4B-Instruct model in Microsoft's research pipeline no longer just visually inspects X-rays, but triggers external tools to calculate dimensions. For threshold-dependent diagnostics, this means the end of merely guessing shapes.

News · 2026-08-10

Model ML uses GPT-5.6 Sol for finance analysis automation

OpenAI highlighted the integration of GPT-5.6 Sol into Model ML, managing finance workflows from data research to traceable PowerPoint decks and Excel workbooks. This marks a shift from raw text generation to producing structured, auditable business documents.

News · 2026-08-10

GPT-5.6-Cyber Moves Red Teaming from the Sandbox to Infrastructure

OpenAI launches GPT-5.6-Cyber as part of the Daybreak Red program. The model is specifically designed for offensive security research and authorized vulnerability testing.

News · 2026-07-18

Reasoning effort becomes a control for LLM cost, time and reliability

Sebastian Raschka explains how LLMs learn low, medium and high reasoning effort modes. For teams deploying models, the practical question is when to pay for longer thinking and when a cheap answer is enough.

News · 2026-08-06

WeatherNext buys cyclones roughly a day of lead time and DeepMind opens the weights

Google DeepMind's Nature paper says WeatherNext improves cyclone track, intensity, and wind-structure forecasts by about a day of lead time. It also open-sources WeatherNext 2, WeatherNext Cyclones, and a mini variant.

News · 2026-08-04

Anthropic moves Tino Cuéllar from its trust into chief global affairs

Mariano-Florentino Cuéllar joins Anthropic as its first Chief Global Affairs Officer. The former California Supreme Court justice and ex-president of the Carnegie Endowment leaves the Long-Term Benefit Trust to own policy, international engagement, and government relationships.

News · 2026-08-04

OpenAI packs Work and Codex into education plugins for teachers and students

OpenAI is shipping three education plugins for ChatGPT Work and Codex: one each for K-12 teachers, college faculty, and college students. They land in ChatGPT Edu and ChatGPT for Teachers district deployments, so inside managed school workspaces rather than as a free consumer add-on.

News · 2026-08-03

Microsoft opens Orchard: a greenhouse for training agents, not another orchestrator

Microsoft Research released Orchard, an open-source framework for scalable agentic model training. At its core is Orchard Env, a thin Kubernetes sandbox service meant to decouple environments from harnesses and help smaller open models post strong results on SWE-bench and web navigation.

News · 2026-08-01

OpenAI solved decade-old math problems for two thousand dollars

The new internal Astra model demonstrated the power of raw compute in mathematics. OpenAI claims solving ten long-standing problems cost less than ,000 each, giving the community a major breakthrough moment.

News · 2026-07-21

Last Week in AI maps the week models, policy and power collided

LWiAI Podcast #252 puts GPT-5.6, Grok 4.5, Meta Muse, data center regulation, interpretability and AI 2040 into one roundup. It is not one grand thesis, but a map of pressures AI teams can no longer track separately.

News · 2026-07-22

Google tested SymptomAI on 13,917 people, but it is not a diagnosis yet

Google Research introduced SymptomAI, a study of conversational agents for symptom interviews and differential diagnosis with 13,917 participants. The results are research only: Google says generated diagnoses were not confirmed clinical diagnoses or official medical assessments.

News · 2026-07-31

Local eval framework shows prompt engineering is moving toward standard developer habits

Simon Willison and Prime Radiant released smevals—a small framework for local model and prompt evaluation. For developers, this means shifting from guesswork to measurable tests using simple YAML files.

News · 2026-07-30

Gemini Robotics ER 2 shifts robots from executors to self-monitoring agents

Google introduced Gemini Robotics ER 2, integrating video understanding and multi-robot orchestration. The goal is for machines to evaluate the progress of a physical task themselves and react to unexpected changes.

News · 2026-07-28

Claude Mythos cut HAWK post-quantum signature strength in half in 60 hours

Anthropic says Claude Mythos Preview found a mathematical weakness in the post-quantum candidate HAWK and improved an attack on reduced-round AES. Neither result breaks production systems today, but both change who can do real cryptanalysis on a weekly cycle.

News · 2026-07-27

Amodei rejects an open-weights ban. He wants chips, distillation limits, and mandatory tests

Dario Amodei wrote that Anthropic has never pushed a ban on open-weights models. Instead of a blanket ban he presses chip export controls, action against industrial-scale distillation, and mandatory safety tests for sufficiently capable open and closed systems.

News · 2026-07-27

Anthropic pushes Claude into enterprise through Cognizant’s army

Cognizant is deploying Claude models into its own tools and training thirty thousand consultants on them. Anthropic gains what a lab cannot build itself: an integrator with corporate compliance clearance.

News · 2026-07-27

AI buying guides moved past chat. They now pick who gets a computer

Ethan Mollick’s summer guide tells most people to pick Claude or ChatGPT from $20 a month and give an agent a computer. Simon Willison frames the shift: the hard choice is now the agent harness and what you allow it to touch.

News · 2026-07-27

After the Hugging Face breach, NVIDIA builds an open AI defense coalition

NVIDIA and dozens of firms including Hugging Face, Microsoft and the Linux Foundation launched the Open Secure AI Alliance. The pitch leans on the July incident where closed APIs slowed forensics and a local open-weight model helped review more than 17,000 actions.

News · 2026-07-21

Gemini 3.6 Flash cuts output tokens by 17%, while Cyber stays restricted

Google has released Gemini 3.6 Flash and 3.5 Flash-Lite for developers, enterprises and the Gemini app. The specialized 3.5 Flash Cyber will be limited to governments and trusted partners in a restricted pilot, giving one model family two very different distribution models.

News · 2026-07-24

Meta AI reaches into calendars and recurring tasks

Meta is adding calendar briefings, event planning, web research and recurring tasks to Meta AI in select markets. The productivity shift makes the chatbot more useful, but it also raises the sensitivity of the data and the cost of every mistake.

News · 2026-07-21

Gemini 3.6 Flash lowers the agent bill while Pro waits offstage

Google released Gemini 3.6 Flash, 3.5 Flash-Lite and 3.5 Flash Cyber with a clear focus on cheaper agentic workflows. For teams paying for tokens and latency, operating economics matter more than the model number.

News · 2026-07-17

Databricks hits a $188B valuation by selling the AI story above enterprise data

Databricks announced new funding at a $188 billion valuation, five months after a $134 billion round. This is not just fundraising: it shows how an enterprise data platform can recast itself as AI infrastructure when it owns the data layer, governance story and cost argument for open-weight models.

News · 2026-07-16

DeepMind turns biosecurity into a program around its AI stack

Google DeepMind and Isomorphic Labs introduced a joint bioresilience approach focused on preventing misuse, detecting outbreaks faster and supporting response to new threats. The notable number is more than 15 partnerships in 12 months, but the harder question is who gets trusted access to these systems.

Library

AI-assisted research — the model as a research partner

AI-assisted research uses models to find hypotheses, write code, test variants and read literature. It is not automatic science. It is a faster research loop with new ways to fall on your face.

Library

Evals and benchmarks — measurement instead of vibes

A benchmark is not truth carved in stone. It is an instrument with error bars. Without it, though, you are only guessing whether a model or agent works.