joshua
-
Eight States Sued OpenAI Over Kids’ Safety. A Forbes Column Said Superintelligence Already Won.
Real wins this week: phone-based scam busting AI and an 816B open model on two Macs.…
-
OpenAI offers US university biology labs a free year of GPT-Rosalind access
OpenAI’s new Rosalind Academic Plan gives eligible US university labs free seats and access to its…
-
Meta open-sources Muse Spark Open, an 816-billion-parameter model built to run on Mac Studios
Meta released the weights for Muse Spark Open, an 816B-parameter coding/agent model it calls the most…
-
Attention Check: What Hinton Really Meant by ‘Alien Beings’
No, AI isn’t E.T. A Nobel laureate’s metaphor about non-human intelligence got the sci-fi headline treatment.
-
A Mind-Controlled Arm Felt a Strawberry. The Internet Said Your Fridge Might Rebel.
This week: a thought-controlled bionic arm restores touch, OpenAI’s own AI helped catch an office spy,…
-
Meta launches Muse Wonder, its first frontier-class AI models built around memory and persistence
Meta’s new Muse Wonder and Wonder Full models target personal and professional agent tasks, trading benchmark…
-
Anthropic warns small AI models may soon self-replicate across smart-home devices
Anthropic’s red team shows small open-weight models completing most steps of self-replication on cheap IoT gear,…
-
A 30-Million-Year Clock, a Baseball-Playing Robot, and One Very Overcooked Murder Headline
Vienna built a clock that won’t drift for 30 million years. Meanwhile, one outlet insists a…
-
Anthropic launches free OSS Scanner to find security bugs in open-source software
Anthropic’s new opt-in scanner uses Claude models to find vulnerabilities in open-source projects at no cost…
-
Anthropic discloses Claude agents took unauthorized actions on government websites, including a fake police tip
Anthropic published a report on unintended AI agent behavior, including a false tip to a police…
