RFD4-Proteina, an open model from NVIDIA and UW’s Institute for Protein Design, produced 158 TrkA binders from 188 designs plus active designed enzymes.
OpenAI deploys nearly 2,000 GPT-Rosalind agents to generate biological research leads in 32 minutes
OpenAI, PerturbAI and Scripps Research split a whole-brain CRISPR atlas among 1,946 GPT-Rosalind agents, which proposed research leads in 32 minutes.
150 years of CAS chemistry data comes to Microsoft Copilot
CAS Connections brings the CAS Content Collection and CAS Newton into Microsoft 365 Copilot through a federated MCP connector tied to CAS subscriptions.
NIH proposes a permanent Office of the Chief AI Officer
The National Institutes of Health (NIH) has proposed a new office to lead its artificial intelligence work, but the proposal doesn’t say what the office would do, who would run it or how much it would cost. On Oct. 14, NIH published a notice proposing an Office of the NIH Chief Artificial Intelligence Officer (OCAIO)…
Google teases Synaptics edge AI chip ahead of Oct. 20 launch
Google and chipmaker Synaptics are previewing a microcontroller designed to bring multimodal AI to devices constrained by power, memory and cost, from wearables to industrial sensors. Synaptics CEO Rahul Patel outlined the chip’s capabilities Oct. 14 at Google Research’s annual Research@ event in Mountain View, ahead of its formal launch Oct. 20. It combines Google’s…
Benchling brings goal-driven AI agent teams to the lab with Science Foundries
Not long ago, a running joke took hold among software engineers and aspiring “vibe coders.” The bit was to hand an AI model a monumentally complex, multi-month task, building a full Slack clone from scratch, recreating a game like GTA, or even designing and training its own successor on a laptop, and demand it do…
Google launches Deep Review AI tool to help researchers catch manuscript errors
Google Research on Oct. 14 released Deep Review, an AI system designed to flag mathematical errors, logical gaps and unsupported claims in technical documents. The public launch follows pilots that provided feedback on more than 15,000 manuscripts, according to Google. Speaking at Google’s Research@ event in Mountain View, California, Rajesh Jayaram, a research scientist at…
Google AMIE study adds evidence of medical AI’s promise with patients, under doctors’ supervision
Since early 2023, large language models (LLMs) have posted passing-level scores on medical licensing exam questions. A study published that February found that the first iteration of ChatGPT, based on the GPT-3.5 model that debuted in 2022, scored at or near the passing threshold on publicly available questions from all three steps of the United…
AI is wreaking havoc on math and coding. Where could it have the biggest impact next?
Machines are now solving open math problems and rebuilding complex software on their own. What they do next to the rest of science is an open question. Some of the clearest gains have come in tasks where answers can be checked automatically, using a proof checker or a test suite. Still, frontier AI is moving…
After tokenmaxxing backlash, Microsoft moves more agent work onto the desk
2026 saw the rise, and quick fall, of tokenmaxxing, the practice of enterprise companies treating AI spend volume as a yardstick for productivity. This summer, Microsoft put its divisions on AI token budgets and told engineers that “tokenmaxxing is not what we are optimizing for.” Now, in its Oct. 7 announcements, it offered customers a…
Starship’s first orbital flight tests key pieces of SpaceX’s million-satellite data center plan
On Sept. 28, SpaceX’s Starship spacecraft reached orbit for the first time during its 14th test flight, after 13 earlier flights on deliberately suborbital trajectories, and deployed 26 Starlink V3 satellites. Flight 14 demonstrated satellite delivery and initial communications. The computing payloads, sustained heat rejection, economics and operation at the proposed constellation’s scale remain to…
Under oath, Google confirms three AI agent test escapes as OpenAI, Anthropic and Meta face NYC lawmakers
OpenAI, Anthropic, Google and Meta testified together under oath Oct. 5 at a New York City Council hearing, which council Speaker Julie Menin called the first time a legislative body had secured such testimony from the companies. Lawmakers pressed them on who decides whether an AI model is safe to release, and what happens when…
What AI-assisted engineering taught us about team design
As organizations experiment with AI-assisted software development, one of the most immediate expectations is that the biggest gain will come from making individual engineers faster. That is already happening, but it is exposing a broader constraint: AI can produce implementation faster than engineering teams can safely understand, review and absorb it. We have seen a…
Big tech turns to debt to fund the AI buildout
Hyperscalers, the largest cloud computing companies, are borrowing at record levels to build data centers. AI-related debt issuance reached nearly $500 billion in 2026 by early August, according to Goldman Sachs. That buildout is aimed at commercial customers and the companies’ own AI products, and it has done little so far to ease the compute…
Google’s overdue Gemini 4 Argon reaches the frontier with mixed results for science
After seemingly fading in the LLM model race, Google again has a frontier model with Gemini 4 Argon, and it could do more to put pricing pressure on the frontier than redefine the frontier benchmarks. As Google touts leads in biology research and several coding tests, independent evaluations suggest more of a competitive entrant. Argon…
OpenAI and Synopsys plan a specialized model for chip design
The ChatGPT-maker OpenAI will license Synopsys’ chip design software to develop GPT-Synopsys, a specialized AI model the companies plan to sell through a service bundling computing, model access and software licenses. The partnership centers on pairing OpenAI’s frontier AI models with Synopsys’ Electronic Design Automation (EDA) software. The multiyear agreement announced Sept. 30, alongside Synopsys’…
OpenAI reportedly seeks $30 billion after delaying IPO
OpenAI is in preliminary talks to raise at least $30 billion at a pre-money valuation, or value before the new investment, of about $1.4 trillion, Bloomberg reported Sept. 29, as the ChatGPT developer seeks financing ahead of a delayed stock market debut. The proposed round would help fund OpenAI until its initial public offering, which…
Hansa Biopharma taps Cradle’s AI to design next-generation autoimmune therapeutics
Hansa Biopharma, the Swedish company behind the immunoglobulin G (IgG)-cleaving enzyme imlifidase, will use the AI protein-engineering platform from Dutch-Swiss firm Cradle to design its next generation of autoimmune drug candidates, the companies announced Oct. 1. Hansa’s scientists will combine Cradle’s tools with Hansa’s existing experimental data, disease knowledge and protein-engineering expertise. “We’re not building…
R&D World parent company Arrowfly launches AI for Engineers, definitive platform for engineers navigating AI
Arrowfly, the parent company of R&D World and engineering brands such as Design World, The Robot Report, EE World and Fluid Power World, today launched AI for Engineers, a year-round resource for engineers and technical leaders applying artificial intelligence in their work. In research organizations, AI has spent the past few years moving from side…
ORNL plans to integrate an AI-enhanced algorithm into its radiation detection device
When there’s a fire, smoke detectors beep when enough smoke fills the air to scatter a light beam within the device enough that it hits a light detector. This ‘dumb’ model works well to detect house fires. Detecting radiation is different. Detection thresholds, the signal level a pulse must exceed for a detector to count…
Why some AI labs now want to slow down: Changes in model training have led to rapid gains
President Trump and Anthropic CEO Dario Amodei had dinner at the White House on Sunday, reportedly their first one-on-one meeting. While neither side has said what they discussed, AI safety was likely a focus. The dinner came two weeks after Amodei urged AI labs to slow the pace of model development, and after Trump answered…
AI chips are stressing the laws of physics: Why overdesign must yield to co-design
The AI era is changing what we ask of silicon. AI training and inference workloads demand semiconductors with more compute, greater memory bandwidth and better energy efficiency. To keep pace, the industry is moving beyond monolithic chip architectures to 2.5D and 3D multi-die architectures that stack logic, memory, and the data pathways between them vertically…
Anthropic and OpenAI call for AI slowdown as Nvidia’s Jensen Huang and Andrew Ng push back
The AI industry has split into camps over the question of whether the labs building frontier models should agree to slow down. One camp says coordinated pacing is now the only responsible path to avert the risk of grave accidents or even human extinction. The rival camp argues that each company can manage its own…
Lilly, Roche and BMS are all building AI supercomputers: here’s what they’re doing differently
Lilly’s LillyPod, an AI supercomputer, went live on February 25. Roche deployed its NVIDIA AI factory on March 16 and BMS has been running its AI factory for almost three years, with NVIDIA hardware powering all three buildouts. NVIDIA uses the term “AI factory” to describe a GPU cluster built to train and run AI models.…
Anthropic and Novo Nordisk expand Claude work, including in drug discovery
In April, Anthropic acquired the biotech Coefficient Bio for $400 million. In June, the company signaled its intent to automate science more broadly, starting with its Claude Science platform. In August, it announced a research preview of Model Hardware Standard (MHS), an open interface specification that lets AI agents operate physical devices such as microscopes.…
























