AI & Robotics

Google says teams used Gemini and other AI tools to help create films, visuals, and event elements for Google I/O 2026, presenting the conference as an internal showcase for AI-assisted production.

OpenAI has rebuilt its robotics team, starting with infrastructure tasks while CEO Sam Altman describes a longer-term goal of personal robots for everyone.

An Anthropic study found large disparities in coding-agent adoption across social science, with gaps by gender, field, career stage, and university rank.

Anthropic reportedly prohibits AI assistance in live job interviews unless explicitly allowed, as the company tries to evaluate how applicants reason on their own.

New research from Kaikaku.AI argues that food recommendation systems should distinguish between ingredients that appear together in recipes and those that are chemically similar.

MISUMI has launched MISUMI Americas as part of a $1 billion investment plan, pairing its precision parts business with Fictiv’s AI-powered digital manufacturing platform.

Mathematician Terence Tao argues AI and formal verification could finally let mathematics adopt a real division of labor, provided automation is paired with rigorous checking.

Reports suggest Nvidia-based Windows PCs from Dell and Microsoft are coming, alongside software that lets AI agents handle tasks locally on the device.

A large behavioral dataset study found that base language models predicted human responses better than post-trained assistant versions across several model families.

NIST has proposed a standardized, low-footprint benchmark for humanoid locomotion and manipulation, aiming to create a common way to evaluate commercially relevant robot capabilities.

Amazon has reportedly shut down an internal AI ranking dashboard after workers inflated their standing with low-value AI activity that raised cloud costs instead of productivity.

A new review argues that autonomous AI depends less on raw model capability than on the software harness that gives models tools, memory, testing, and execution loops.

OpenAI says it will provide access to its life sciences model GPT-Rosalind for vetted public-interest biodefense and pandemic-preparedness projects.

OpenAI says GPT-5.5 Instant is being updated for more natural, readable responses while older models begin their sunset inside ChatGPT.

Google’s partnership with the University of Waterloo is producing working AI prototypes aimed at education, accessibility, and coaching rather than just classroom theory.

The final day of the 2026 Robotics Summit & Expo highlights open robotics, AI productization, telesurgery latency and Neuralink-enabled human-computer interaction.

Meta is rolling out paid add-ons across Instagram, Facebook and WhatsApp while building separate AI subscription tiers, signaling a broader attempt to monetize AI compute and reduce reliance on advertising.

Mistral has rebranded Le Chat as Vibe, bundling chat, workplace integrations, coding agents, and recurring workflows into a broader product aimed at competing with enterprise AI agent platforms.

Google has unveiled a compact Coral Board that can run Gemma 3 locally, signaling a stronger push toward on-device AI for wearables and other small hardware.

Google Cloud has introduced AI Threat Defense, a platform that combines Gemini, Wiz, DeepMind’s Codemender, and Mandiant tools to find exploitable flaws and generate tested fixes faster.