Google Labs · Original source
Google Labs has introduced Tabletop, an experiment for real-time group collaboration. Users add links, files and ideas, then prompt @tabletop or pick a starter template to generate interactive trip itineraries, dinner polls, expense splitters and tournament brackets, and invite others to join the space. Tabletop is rolling out in the U.S. to users aged 18 and over through a waitlist at labs.google/tabletop.
ElevenLabs · Original source
At its New York Summit, ElevenLabs announced plans to help businesses fine-tune open-source language models and its own voice models, previewed an Interaction Engine, and described ElevenAgents upgrades including Memory and Routines. It previewed ElevenAI, an app for dictation and meeting notes built on Scribe v3, and launched a Partner Program covering more than 550 partners in 44 countries, an alliance with McKinsey and the ElevenLabs Marketplace.
ElevenLabs and NVIDIA · Original source
ElevenLabs and NVIDIA have expanded their partnership. NVIDIA Nemotron 3.5 Super VL is now available in ElevenAgents, and ElevenLabs is adding 15 MW of compute based on NVIDIA Blackwell Ultra and Vera Rubin platforms. ElevenLabs also powers a sovereign voice agent for Poland's national e-Health platform appointment system, running on-premises with NVIDIA Confidential Computing, and is exploring on-device voice AI for robotics.
Salesforce · Original source
Salesforce AI Research has published SFR-AutoR&D, a set of agents that run the research loop end to end to improve codebases and train AI models. Applied to its SFR-Embedding V3 retrieval model, it raised the nDCG@10 score by 6.13 points on 10 MTEB tasks, then in a second session of about nine hours delivered 36% more throughput and 17% lower latency. Salesforce says researchers still set the goals, budget, data constraints and evaluation.
OpenAI · Original source
OpenAI signed the Charter of Engagement for Safer AI for Children at the Paris Peace Forum and committed $2 million to the scientific research network under the Coalition for Children’s Rights and Protection in the Age of AI. The funding supports shared measurement tools for assessing AI's risks and opportunities for children. The charter provides for research independence, with findings published including negative or inconclusive results.
Google · Original source
Google is bringing the Gemini app to Apple CarPlay, letting iPhone users hold natural voice conversations while driving using Gemini Live. Drivers can ask about their calendar, catch up on email through Google Workspace connections, find local recommendations or rehearse a presentation. Chats sync with the Gemini app on the phone. The rollout is global, starting on 11 November, for Gemini users on iOS across every supported Gemini Live language.
Google DeepMind · Original source
Google has introduced the Gemini Decisions API, powered by a new model, Gemini Decision 1, for fast closed-set decisions such as tool selection, policy checks and model routing. One multimodal input of up to 256K tokens can be evaluated against up to 1,000 yes/no, choice or rubric-scoring questions in a single pass, each with a confidence score. It is priced at $0.10 per million input tokens with no output-token charge.
IBM · Original source
IBM's report Proof, Not Promises: The Digital Sovereignty Index finds an overconfidence gap among large enterprises. Of those surveyed, 94% faced at least one sovereignty challenge in three years and 81% of them had to give formal evidence of control. An average of 82% are confident they can prove control, but only 26% reached the Index's Leading tier. AI-related issues were the most common challenge, at 44%.
OpenAI · Original source
OpenAI says more than 20,000 transitioning servicemembers and veterans have claimed a free year of ChatGPT Plus in the past year. OpenAI Academy and community partners have run more than a dozen trainings for nearly 1,000 members of the military community, with seven more scheduled before the end of the year. The post was published ahead of Veterans Day and describes workshops in Reston, Virginia and Louisville, Kentucky.
Google DeepMind · Original source
Google DeepMind has announced AlphaGenome 2, a model that predicts how DNA sequences and variants affect specific human cell types. DeepMind says it identifies nearly twice as many variants that change gene expression in cells as its previous state-of-the-art model, and was trained on more than 100 million individual cell profiles across 68 tissues. It is available for non-commercial research via the AlphaGenome API, with a private preview for commercial research.
