Meta AI News - Page 6 of 17
337 articles • Page 6 of 17
Healthify and OpenAI debut Ria Voice, a realtime AI health coach
Your voice assistant already knows what song you want to hear. But what if it could actually help you lose weight?
Meta's Muse AI model now generates images for Instagram, WhatsApp
Meta rolled out its first image generator built entirely inside Superintelligence Labs on Tuesday, and it's already live in the Meta AI app,...
SMG releases smg-grpc-proto on PyPI; vLLM integrates via PR #36169
Multimodal AI is slow for a boring reason. Python is in the way. Before any image reaches the brains of a model like Llama 4 Vision, it gets chewed...
Amazon adds agentic fine‑tuning to SageMaker for Llama, Qwen, Deepseek, Nova
Fine-tuning an AI model has been grueling, specialist work. Now it's over. Amazon declared it so Wednesday with a new agent for SageMaker, its cloud...
Meta researchers find signatures in LLM traces signal reasoning correctness
There’s a basic problem with the machines we’re told will reason. We have no real idea if they are.
New Magnetic Nanoparticle Approach Merges Heating and Healing for Bone Cancer
Bone cancer treatment is brutal. Surgeons and oncologists burn tumors out, a process that too often also incinerates a patient's chance at a normal...
Alabama AG Subpoenas OpenAI Over Safety Practices Following Hack
Alabama Attorney General Steve Marshall issued a subpoena to OpenAI on Monday, opening a formal investigation into an incident last month where one...
Prime Intellect's Open-Source Agent Runs Sub-Agents as IPython Kernel Functions
Prime Intellect released Prime Agent on Wednesday, an open-source coding harness licensed under MIT that throws out two of the standard assumptions...
DRAM Shortages Lift Nvidia AI Server Prices by 15 Percent
Servers built around Nvidia's Vera Rubin and Grace Blackwell chips are getting more expensive, and by a lot.
Sarah says plugging memory into a harness is like plugging driving into a car
You can’t just bolt memory onto an agent like a spare part. Sarah makes that brutally clear: asking to plug memory into a harness is like asking to...
Meta AI Launches Muse Code Beta, Powered by Muse Spark 1.2
Meta AI put out a beta version of Muse Code on Tuesday, a terminal-based coding agent built to run on Muse Spark 1.2, a model the company trained...
Anthropic CEO Says AI Centralizes Power, Open Models Shift It to Chip Owners
Dario Amodei has spent the past week defending himself on X against a charge that his company's regulatory push is really a power grab dressed up as...
eOptShrinkQ enables near‑lossless KV cache compression with spectral denoising
Compressing the key-value cache without destroying signal fidelity has long been a tug-of-war between quantization error and the noise that...
Arcee launches open‑source Trinity‑Large‑Thinking as Meta steps back from Llama 4
When Meta’s Llama division stumbled in April 2025, Llama 4 hit by quality complaints and whispers of benchmark manipulation, the open-source frontier...
NVIDIA Outlines AI Infrastructure Advances in Networking and Compute
The data center is no longer a room of humming servers, it’s a factory. And NVIDIA is engineering its blueprint from the ground up.
Siri Joins Gemini: Apple's AI Assistant Enters New Era
Apple has finally admitted Siri isn't smart enough. The company’s solution is to wire Google’s Gemini model directly into its digital assistant.
Time names AI power brokers 2025 Person of Year as firms spend billions
It was the year the brakes came off. Time has named AI’s power brokers its 2025 Person of the Year, a choice that reflects a world where caution gave...
Study finds condensed metaphors and rhythmic framing can evade safety filters
Poetry can break an AI's rules. Not with force, but with form. A new study shows that safety filters, trained to catch obvious harmful requests, are...
Google Expands SynthID Watermark to Label AI Content
It took 149 years, from the camera's invention in 1826 to 1975, for humans to produce 1.5 billion images.
ChatGPT's 'Nerdy' tweak rewards goblin metaphors in answers, study finds
ChatGPT briefly, bizarrely, fell for goblins. The whole fling started with a single broken piece of code.