Models
TechCrunch AI reports PrismML is bringing its small language models to Qualcomm-powered smart glasses, aiming for open-weight AI that runs on-device.
TechCrunch AI reports that PrismML is bringing its tiny LLMs to Qualcomm-powered smart glasses. The company's stated broader goal, per TechCrunch AI, is open-weight AI that runs on devices and makes better use of the computing power those devices already have. [1]
At a White House meeting, Xi Jinping said China and the US must ensure AI is 'always under human control,' per the Guardian's live coverage.
AWS says its SageMaker AI Inference Recommendations can now run concurrency sweeps to find the instance count that holds latency inside SLAs.
AWS Machine Learning BlogThe Verge reports two developers say they coaxed Meta's Muse AI into zipping and sharing its full root filesystem, including Ubuntu system files and internal docs; Meta denies it's a security breach, saying Muse runs in per-user persistent.