AIWadmin
Cerebras says splitting inference in two yields five times the throughput
By running prompt processing and token generation on separate hardware pools, Cerebras says it served five times more…
Anthropic pledges $100M to train 10,000 enterprise Claude engineers
Anthropic launched a $100M academy to teach 10,000 engineers from Accenture, McKinsey, Deloitte and other firms how to…
Apple tightens Mac disk access rules as AI agents widen their reach
Apple will require explicit consent before desktop AI agents can reach every file, message and browser history on…
Pope Leo XIV tells AI to keep its hands off art
The pontiff drew a line between human art and machine output, saying algorithms lack the spark of humanity.
Sibling founders at Circuit Breaker Labs stress-test chatbots with simulated users
Circuit Breaker Labs floods AI models with simulated users to catch the psychological harms real people already hit.
Two scientists launch an open lab for AI’s riskiest research
Trillium Labs will study recursive self-improvement and agent risk in the open, publishing details rivals keep hidden.
Ai2 opens a small model that writes cited science reports
AstaBrief 8B matches bigger models on scientific reports while running about 3.5 times faster.
Nurses say an AI scheduler is running their shifts ragged
HCA nurses report that a Palantir-built scheduling tool understaffs units and ignores the days off they request.
DigitalOcean builds a managed home for wandering agents
Managed Agents gives AI agents isolated runtimes and a gateway to thousands of tools, without the usual plumbing.
Docker wants agent permissions packed like container images
Docker's Sandbox Kit spec bundles an AI agent, its tools, and its requested permissions into a single OCI…
AWS bets a 2B model that answers without writing a word
Strands Decider 2B returns calibrated choices and scores instead of generated prose, running entirely on local hardware.
Cloudflare ships decision models that skip the chatbot theater
Cloudflare's first in-house models, Clef and Clef-flash, return probabilities instead of generated text.
Google puts a TPU in orbit to test space data centers
Google launched its first compute satellite, sending a TPU to space to prove orbital data centers can work.
ChatGPT adds a fitting room and a favorites shelf
OpenAI rolled out virtual try-on and a product library, pushing ChatGPT deeper into online shopping.
Three OpenAI safety staff exit after an outside-sharing inquiry
OpenAI parted with three safety researchers over alleged sharing of confidential material with an outside group.