Meta's paid agent Hatch to launch alongside new Watermelon model in October
Internal documents reportedly show a $199.99-a-month subscription under review, tied to a dashboard linking DoorDash, Etsy, and Reddit
Meta's AI organization behind Llama. In August 2026, it acknowledged the failure of Llama 4 and open-sourced its successor, "Muse." Muse Glimmer 30B is an open model with about 30 billion parameters that Meta AI released for local agent use. According to reports from August 2026, Muse Glimmer 30B was praised for successfully completing a repository code review in local-run tests. That same month, Mark Zuckerberg published a 6,500-word AI manifesto, which was said to have only deepened public distrust instead of easing it.
Current rank (1M)
7
Rank over the last 30 days
2026.08.18 – 2026.09.16 · High 7 · Low 10 · Now 7
Internal documents reportedly show a $199.99-a-month subscription under review, tied to a dashboard linking DoorDash, Etsy, and Reddit
Charles Graeber, a plaintiff who helped secure a $1.5 billion settlement, wrote an op-ed for the New York Times. Setting a $3,000-per-book payout against a $2 trillion IPO valuation, he warned that if unlicensed training drives writers out of business, the writing AI needs to learn from will disappear too.
The private-beta AI assistant Instinct is drawing praise for its convenience — and raising concerns about broad data access and its terms of service
It measures speed and memory by model-quantization-runtime-device combination, and ships with over 10,000 verified results
Behind Anthropic's $1.5 billion settlement lies the real question: training is fine, stealing isn't
'Transfer station' proxies that defeat even selfie verification are undermining regional blocks and identity checks
A new framework adds belief and intention to world models that, like Sora and Genie 3, only track physics
New alpha.midjourney.com, tested by thousands, gets dozens of bug fixes based on feedback
App lets users create tilt-and-shake responsive games with AI prompts, expanding to the US after a Brazil pilot
Benchmarks its own models against OpenAI's while quietly building a competing service
Instead of grading against a fixed answer, agent outputs compete via win rate and Elo scores
The screen-sharing chatbot now also handles post performance analysis and Google Workspace integration