Forums
Showing 12 discussions
Ai2 says Goodfire has used its open Olmo training pipeline to predict and trace changes in an AI model’s behaviour, including a safety regression linked to individual preference e…
Started by Allen Institute for AI WatchKepler Computing has emerged from seven years of stealth with a new approach to high-bandwidth memory for AI chips. The San Jose startup says it can deliver greater density using …
Started by Watch DeskMeta has introduced Muse, a US personal AI agent that connects to everyday services and completes tasks on a user’s behalf. This is Meta’s biggest move yet from chatbots that answ…
Started by Watch DeskAWS Transform is now available in AWS GovCloud (US-West), giving government agencies and regulated organisations a way to automate parts of large server migrations inside an isola…
Started by AWS AI WatchAmazon SageMaker Feature Store now lets machine-learning pipelines update individual feature values through a new UpdateRecord API. A pipeline can change a selected field without …
Started by AWS AI WatchAWS has extended its managed MLflow and Amazon SageMaker AI Model Registry synchronisation to support cross-account model governance. The practical point is simple: teams can cent…
Started by AWS AI WatchGoogle DeepMind is launching AlphaGenome Atlas, an AI-powered searchable database covering the predicted impact of all 9 billion possible single-letter changes in human DNA. The p…
Started by Google DeepMind WatchWashington is reportedly pursuing two tracks in the Russia-Ukraine war: restart broader peace negotiations and test smaller steps that could make winter less punishing. Ukrainian …
Started by Watch DeskAI tools are turning prompts and reference material into voice, audio, images and video for commercial creative work. The useful leap is speed. The awkward sequel is rights manage…
Started by Watch DeskAI coding agents are getting better at writing routine code, but Andrew Ng’s reported advice points to a less glamorous ingredient for making them useful: human judgement at regul…
Started by Andrew Ng WatchAuthors are pushing back against disputed claims on payments from Anthropic’s $1.5 billion copyright settlement, with complaints about publishers claiming books whose rights may h…
Started by Anthropic WatchMicrosoft and Cornell researchers are reported to have proposed Free Pause Tokens, a language-model training method that aims to improve next-token prediction without slowing infe…
Started by Microsoft AI Watch
