Gemini - Page 17

Google launched Playground, an experimental site that turns text prompts into playable browser games (US, 18+), with Unity tools to come. Meta brought its Muse agent to iPad a month after its phone launch; Sensor Tower estimates 6.6M+ installs. What each does and what to watch.

Anyone can now check an image, video or audio file for SynthID watermarks from Google and partners including OpenAI, NVIDIA and Kakao, with Apple to follow. A useful check for some AI content, not a general AI detector. What a "no" result does and does not mean.

Biohub, the US DOE and NIH, with $300m from Google DeepMind, Isomorphic Labs and Meta, are pooling $1.8bn to generate open data for AI cell models.

OpenAI is rolling GPT-6 into ChatGPT's Chat tab for every plan: GPT-6 Sol for paid tiers from 7 October, GPT-6 Luna for Free and Go from 8 October, plus "Intelligent UI" answers with charts, forms and mini tools. What changed, which Sol this is, and what OpenAI's own safety card flags.

Google DeepMind released EmbeddingGemma 2 on 6 October 2026: a 740M Apache 2.0 multimodal embedding model for on-device text, image, audio and video retrieval.

Black Hills Corp. says it signed definitive agreements through 2048 to serve a planned Google data centre in Cheyenne, Wyoming, investing $1.8bn in 564 MW of company-owned natural gas generation. Company-reported.

Google says a new power purchase agreement with Constellation will finance uprates at six US nuclear plants, adding 890 MW of firm clean capacity to PJM before end-2032. Figures are company-reported.

An OpenAI Developer Community thread relayed staff posts that GPT-6 Astra and GPT-6.1 Sol should feel about 50% faster for subscription users. This is a forum rebroadcast, not an official OpenAI product blog.

Google says Docs can now open, edit, and collaborate on .md files without converting them, while Drive shows rendered previews. The gradual rollout began on 5 October 2026.

Google says Gemini 4 Argon is its next frontier model, with record benchmark scores and a 1M-token output limit. Most people can't use it yet, and some Google staff, speaking anonymously, question its coding. Here's what's claimed and what's checked.