DevFest 2026 Highlights Agentic AI Across 800+ Events
Google's DevFest 2026 returns with more than 800 global events focused on building, securing, and scaling in the agentic AI era.
AI & ML
Frontier models, research results, agents, and developer tooling.
Perplexity Deploys GPT-6 Astra for End-to-End System Work
OpenAI says Perplexity now trusts GPT-6 Astra to handle communications, software changes, and production monitoring with less human oversight.
How Fyxer uses OpenAI models to build a trusted AI email assistant
A case study from OpenAI shows how fine-tuning, memory, and user feedback help Fyxer draft emails that sound like each person.
Unitree Robot Dog Review: $4,000 Quadruped Impresses
A hands-on review of Unitree's $4,000 robot dog suggests the Chinese company may be the world's most important robotics firm.
Forward Deployed Engineers: Best Practices for AI Deployment
A new playbook from Palantir and Kepler shows how forward deployed engineers turn AI models into working business outcomes.
How The Den Uses ChatGPT Work to Save 10-15 Hours a Week
A family social club cuts administrative work from days to hours with OpenAI's ChatGPT Work as it expands.
DeepSeek v4.1-Flash: 763B Causal Encoder-Decoder With Vision
DeepSeek's v4.1-Flash introduces a large causal encoder-decoder with vision, a shift significant enough that Latent Space says it should have been called v5.
New Mexico lawyer penalized for using ChatGPT's invented witnesses
A defense attorney was sanctioned after court filings contained fabricated testimony produced by ChatGPT, highlighting the risks of unverified AI use in legal work.
AI Ataraxos Defeats Top Stratego Player Using Belief Model
A new AI called Ataraxos has beaten the world's best Stratego player, solving a long-standing challenge in imperfect-information games with a belief model and limited compute.
Why AI Might Matter Most for Monotonous Work
An OpenAI-hosted essay argues that AI will accelerate progress by easing the execution bottleneck, not just by generating brilliant ideas.
Gemini 4 Argon Arrives With 1M-Token Output, Limited to Cyber Testers
Google DeepMind's new frontier model posts top benchmark scores and a record output limit, but initial access is restricted to a cybersecurity preview.
Olmo-core 3: open training stack for trillion-parameter MoEs
Ai2's open-source Olmo-core 3 redesigns mixture-of-experts training to keep throughput high as models scale past a trillion parameters.
GeForce NOW Adds 25 Games in October, Including Witcher 3 Remastered
NVIDIA's cloud gaming service expands its October lineup with 25 titles, though the announcement doesn't touch on AI or machine learning.
AI Leaders Urge Brakes After Reckless Speed, Citing Safety
After years of rapid deployment, AI leaders now emphasize caution, a shift that may also protect established industry players.
Christina Koch and Google's James Manyika talk space and AI
Google is pairing an astronaut with its head of research to frame space exploration as part of its AI-focused public conversation.
Children’s Hospital Uses Open Source NVIDIA AI for Cardiac Care
A major children’s hospital is applying open source NVIDIA AI tools to cardiac care, according to a vendor blog post.
Google Opens AI & Economy ATLAS to the Public
The interactive, open-access platform turns millions of global data points into a public window on AI’s economic impact.
Google Moves Beyond Translation to Full Language Understanding
The company says its AI efforts now target modeling the world's living languages as they are truly used, not just converting text between them.