DeepX claims inference up to 100B parameters at less than 5 W
We haven't written up this one. Hacker News has the full story — the link below goes straight to it.
More in AI
Jensen Huang disappears from lineup at Anthropic and ElevenLabs events
Jensen Huang is no longer on the agenda for an Anthropic forum, according to a person familiar. He also vanished from ElevenLab's conference lineup.
Meta expands AI age detection globally to protect teens online
Meta has announced the global expansion of its artificial intelligence (AI) age-estimation technology to identify teenagers who misrepresent their ages online and automatically place them under its…
Apple Continues to Say Some AI Features Will Have Fees for Extra Usage
Apple continues to advise customers that certain Apple Intelligence features that rely on its server-side Private Cloud Compute system have daily usage limits, with expanded access to such features to…
Google claims its new AI software cuts inference latency by over 90% — but there’s a catch
Google’s eighth-generation TPU announcement on Monday included more than new chips, introducing TPU Sync, an open-source system designed to speed The post Google claims its new AI software cuts…
AI use is now universal in higher education. Why are universities failing to adapt?
Unis must engage deeply with the discussion about how to adapt to artificial intelligence. Otherwise they risk becoming redundant.
Most uni students now use AI, but that doesn’t mean they like it
We asked more than 10,000 students about how they use artificial intelligence for study and how they feel about the technology. This is what we found.