Search “Google”

Google DeepMind expands PlanAssist to 12 more local planning authorities in England

The AI planning tool PlanAssist, developed by Google DeepMind in partnership with the UK government, will be expanded to 12 additional local planning authorities in England, bringing the total number of participants to 15. In early pilots in Barnet, Camden and Dorset, the tool cut the average time planning officers spent assessing household residential applications by 58%, and more than 70% of the generated reports were approved by officers without modification. Built on Gemini, PlanAssist verifies application materials and drawings, checks against heritage protection restrictions and local and national planning policy, reviews past objections, and drafts reports with citations, with the final decision still made by planning officers. The tool has been piloted since June and is now entering a 12-month Beta phase; if testing goes smoothly, the UK government plans to make it available to all local councils in England in 2027.

Google DeepMind expands PlanAssist to 12 more local planning authorities

The AI planning tool PlanAssist, developed by Google DeepMind in partnership with the UK government, will be expanded to 12 additional local planning authorities in England, bringing the total number of participants to 15. In early pilots in Barnet, Camden and Dorset, the tool cut the average time planning officers spent assessing household residential applications by 58%, and more than 70% of the generated reports were approved by officers without modification. Built on Gemini, PlanAssist verifies application materials and drawings, checks against heritage protection restrictions and local and national planning policy, reviews past objections, and drafts reports with citations, with the final decision still made by planning officers. The tool has been piloted since June and is now entering a 12-month Beta phase; if testing goes smoothly, the UK government plans to make it available to all local councils in England in 2027.

View story:Google DeepMind expands PlanAssist to 12 more local planning authorities in England

Google receives the first AI response from the Project Suncatcher prototype satellite

On October 30, 2026, Google said it had received the first AI reply from the Project Suncatcher prototype satellite. The data-center-class TPU aboard the satellite ran the open Gemma 4 model in orbit; after the ground station uploaded a prompt, the model generated a haiku about space and transmitted it back to Earth as the satellite passed over Svalbard, Norway. The satellite launched on October 1 aboard SpaceX Transporter-18 and was built in partnership with Planet; it had previously completed power-on of the host machine and TPU tray as well as early machine learning tests. Google will next test how the TPU performs under orbital thermal stress and radiation exposure, paving the way for the goal of two interconnected satellites in 2027.

Google Developer Groups announced that DevFest 2026 will be held from October 1 to December 31

Google Developer Groups announced that DevFest 2026 will be held from October 1 to December 31. According to information published on the Google Developers Blog, this year's DevFest covers 115 countries and more than 800 events, with nearly 1 million developers expected to participate. The events revolve around generative AI, autonomous agents, Google Cloud, Android, and the Web, under the theme of building, securing, and scaling in the age of agents. Local events are organized by communities in local languages, with Google Developer Experts and Google engineers participating as speakers.

Google receives first AI response from Project Suncatcher prototype satellite

On October 30, 2026, Google said it had received the first AI reply from the Project Suncatcher prototype satellite. The data-center-class TPU aboard the satellite ran the open Gemma 4 model in orbit; after the ground station uploaded a prompt, the model generated a haiku about space and transmitted it back to Earth as the satellite passed over Svalbard, Norway. The satellite launched on October 1 aboard SpaceX Transporter-18 and was built in partnership with Planet; it had previously completed power-on of the host machine and TPU tray as well as early machine learning tests. Google will next test how the TPU performs under orbital thermal stress and radiation exposure, paving the way for the goal of two interconnected satellites in 2027.

View story:Google receives the first AI response from the Project Suncatcher prototype satellite

Google announces that DevFest 2026 will be held from October 1 to December 31

Google Developer Groups announced that DevFest 2026 will be held from October 1 to December 31. According to information published on the Google Developers Blog, this year's DevFest covers 115 countries and more than 800 events, with nearly 1 million developers expected to participate. The events revolve around generative AI, autonomous agents, Google Cloud, Android, and the Web, under the theme of building, securing, and scaling in the age of agents. Local events are organized by communities in local languages, with Google Developer Experts and Google engineers participating as speakers.

View story:Google Developer Groups announced that DevFest 2026 will be held from October 1 to December 31

Google Research releases the SLEEP offline memory consolidation method

Google Research releases the SLEEP offline memory consolidation method. According to a report by IT Home · AI, Google's research team has released a continual learning study called the "Sleep Framework," which divides model operation into two phases, awake and sleep, to address the problem of large models having fixed knowledge and difficulty absorbing new information; during the sleep phase, short-term context is written into long-term parameters through parameter expansion, mixture-of-experts memory blocks, and knowledge distillation, and synthetic samples generated through "dreaming" are used for reinforcement learning. Experiments show that the framework can handle contexts of up to about 10 million tokens, that performance degradation is close to zero in multi-stage continual learning on unseen languages, and that it achieves higher accuracy on math benchmarks such as AIME 24/25. Phoenix Technology reported that the study was made public on October 27, and that the paper is titled "Language Models Need Sleep: Learning to Self-Modify and Consolidate Memories." There are currently no subsequent developments.

Google releases the Gemini 4 Carbon model

On October 28, 2026, Google DeepMind released Gemini 4 Carbon, calling it its strongest model, ranking first with 59 points on the Artificial Analysis Intelligence Index v5.0, and scoring 88.2% on SEC-bench Pro, 54.8% on ExploitGym, and 71.0% on CWE-bench. The model is aimed at long-horizon engineering tasks and was first opened to some cybersecurity defenders before being expanded through the Fairwind Program. On October 29, Arena.ai said Gemini 4 Carbon ranked 2nd on Agent Arena with a net improvement of 16.2% and pushed the Pareto frontier forward; the same day, Yuchen Jin cited Artificial Analysis data to raise reservations, noting that the evaluation used the Antigravity CLI, and he planned to try it instead in Codex and Omnigent. On October 30, the Artificial Analysis Intelligence Index showed Carbon in first place with 59 points, 1 point higher than Claude Opus 5.5. On October 31, Perplexity CEO Aravind Srinivas shared an early evaluation, saying it set a new SoTA on the WANDR benchmark. There is also news that

Google Research releases SLEEP offline memory consolidation method

Google Research releases the SLEEP offline memory consolidation method. According to a report by IT Home · AI, Google's research team has released a continual learning study called the "Sleep Framework," which divides model operation into two phases, awake and sleep, to address the problem of large models having fixed knowledge and difficulty absorbing new information; during the sleep phase, short-term context is written into long-term parameters through parameter expansion, mixture-of-experts memory blocks, and knowledge distillation, and synthetic samples generated through "dreaming" are used for reinforcement learning. Experiments show that the framework can handle contexts of up to about 10 million tokens, that performance degradation is close to zero in multi-stage continual learning on unseen languages, and that it achieves higher accuracy on math benchmarks such as AIME 24/25. Phoenix Technology reported that the study was made public on October 27, and that the paper is titled "Language Models Need Sleep: Learning to Self-Modify and Consolidate Memories." There are currently no subsequent developments.

View story:Google Research releases the SLEEP offline memory consolidation method

Google releases Gemini 4 Carbon model

On October 28, 2026, Google DeepMind released Gemini 4 Carbon, calling it its strongest model, ranking first with 59 points on the Artificial Analysis Intelligence Index v5.0, and scoring 88.2% on SEC-bench Pro, 54.8% on ExploitGym, and 71.0% on CWE-bench. The model is aimed at long-horizon engineering tasks and was first opened to some cybersecurity defenders before being expanded through the Fairwind Program. On October 29, Arena.ai said Gemini 4 Carbon ranked 2nd on Agent Arena with a net improvement of 16.2% and pushed the Pareto frontier forward; the same day, Yuchen Jin cited Artificial Analysis data to raise reservations, noting that the evaluation used the Antigravity CLI, and he planned to try it instead in Codex and Omnigent. On October 30, the Artificial Analysis Intelligence Index showed Carbon in first place with 59 points, 1 point higher than Claude Opus 5.5. On October 31, Perplexity CEO Aravind Srinivas shared an early evaluation, saying it set a new SoTA on the WANDR benchmark. There is also news that

View story:Google releases the Gemini 4 Carbon model