Thursday, October 15, 2026
Apple and Meta are working together to let developers run Meta’s Muse Spark Open AI model on multiple Mac Studio computers. The setup runs the model without a cloud service, but it requires several costly machines.
Free daily newsletter
Get these stories in your inbox every morning.
One email a day. No spam. Unsubscribe anytime.
Subscriptions open soon. Check back shortly.
Bee, owned by Amazon, introduced the Frontier wearable and Eclipse phone attachment, which use AI to turn conversations into notes and summaries. The $75 Frontier is available now, while the $100 Eclipse is coming soon; some extra features require an $11.99 monthly Pro plan.
NVIDIA RTX Spark launches Oct. 16 as a Windows PC for creators, gamers and people running AI on their own devices. NVIDIA is marking its release with hands-on events in six cities worldwide.
Meta has released Muse Spark Open, an open-weight AI model built to handle complex coding tasks. It can also work with images, audio, and video, but its size means running it requires substantial computing hardware.
NVIDIA has added three open AI models to its Nemotron family: Super VL for tasks involving text and images, and two Bolt models designed for local devices. Super VL and the larger Bolt model are available now, while the smaller Bolt model is due October 23.
Anthropic is bringing in nonprofits METR and SecureBio to examine its AI risks and safety practices. METR will have broad access to company resources and publish findings, while SecureBio will review part of an earlier risk report.
Bank of America hired 1,000 contractors to fix software flaws that advanced AI tools, including Anthropic's Mythos, can uncover. CEO Brian Moynihan said they joined 3,000 existing staff working on the problem and that repair costs should fall over time.
Bluesky has expanded Attie, its AI-powered tool, so people can make websites, games, polls and other interactive projects. The creations can connect to users’ Bluesky identities and reach people on its decentralized network.
The FTC has requested records from Anthropic and OpenAI as it investigates possible consumer harm linked to AI safety incidents. The requests cover the companies’ security measures and marketing claims, and investigators may also seek information from others affected by their AI agents.
Anthropic has given six METR staff broad access to its internal systems to assess its AI models and safety practices as independent evaluators. METR can publish its findings without Anthropic’s approval, while other reviewers currently have less access.
Free daily newsletter
Join readers who get Training Data Daily before their first meeting.
One email a day. No spam. Unsubscribe anytime.
Subscriptions open soon. Check back shortly.