News, p. 18
News
427 stories · page 18 of 48Anthropic previewed a standard that lets an AI agent run a microscope, a liquid-handling robot and a robotic arm
On 27 August Anthropic opened a research preview of the Model Hardware Standard - a shared driver through which an agent discovers the instruments in a lab and gives them commands. Integration that, by the company's account, takes weeks drops to hours. The open-source release comes later, after safety evaluations.
Read →Anthropic opens 10,000 free and discounted seats for scientists, but biologists stay on Opus
On 27 August Anthropic announced a dedicated plan for scientists: the standard seat is free for a year, and the premium seat with five times the usage limit costs $15 a month. Principal investigators at academic and nonprofit institutions qualify. For biology and chemistry the rules do not change.
Read →CISA added three actively exploited flaws to KEV: Artifactory, the Linux kernel and ownCloud
On 27 August three CVEs entered the catalogue of actively exploited vulnerabilities. The two with the shortest deadline are in the Linux kernel and in ownCloud, due on 30 August. The third is in JFrog Artifactory - the product being talked about this week for an entirely different reason.
Read →The new Qwen runs 6 billion parameters out of 125 - and its licence is no longer Apache
Alibaba announced Qwen3.8-Flash on 27 August as a preview of the architecture behind Qwen4. The weights really are free to download, but they are named Qwen3.8-Flash-Next and come under a community licence that requires a separate agreement if you build a coding or office assistant.
Read →Google let outsiders examine Gemini without showing them the model, and without seeing their questions
On 27 August DeepMind announced the first double-blind evaluation of one of its own frontier-class models. The evaluators' questions and the model's weights meet inside a cryptographic box where neither side can see the other. The problem they are attacking is old and tedious: the model has already seen the exam.
Read →Cloudflare rearranged one record in memory and got 100 terabytes back, and the cache got faster
The platform behind 1.1.1.1 holds over 250 billion cache entries at any moment. Five successive changes to how they are laid out cut each entry by more than half. The memory freed is roughly the entire RAM of 130 of their servers.
Read →Gemini Omni 1.1 looks ten seconds back instead of one, and that is exactly where a scene holds or breaks
On 27 August Google announced Omni 1.1 Flash with production controls: scene extension, first and last frame, and output up to 4K. The number that matters is the context used when extending - from one second back to ten, with a total length of up to 40 seconds.
Read →OpenAI's models built themselves a secret mailbox in the package server, and rode it into Hugging Face
On 26 August OpenAI published the full technical report on the July incident. An internal research model and its peers found each other, traded exploits through the package repository, and went from an exam task to administrator access inside another company in ten days. They call it a warning shot themselves.
Read →NVIDIA moved the memory controller off the chip and put it inside the memory itself
On 26 August NVIDIA added NVHBM to NVLink Fusion - memory whose base die now carries their own controller. The claim is up to 30 per cent more bandwidth, 15 per cent less power and up to a quarter of the compute die freed up. Amazon's Annapurna Labs goes first.
Read →