OpenAI Releases GPT-5.5, Its Fastest Model Yet for 'Intelligence-Bottlenecked' Work

DJI launches its new beginner drone series with 48MP cameras, omnidirectional sensing, and up to 52 minutes of flight time. The catch: no US availability.

Half a million Britons' genetic and health records appeared on Alibaba. The breach underscores a growing vulnerability in centralized medical data systems.

The new surf-focused G-SHOCK pairs tide graphs and moon data with optical sensors and Polar-powered training analytics.

Panasonic revives its travel zoom line with a backlit sensor upgrade and USB-C, though the EVF didn't survive the transition.

DJI applies its aerial navigation expertise to home cleaning with impressive results, though a recent security breach and US restrictions complicate the picture.

Google unveils two distinct TPUs at Cloud Next 2026, with the 8t targeting frontier model training and 8i optimized for the agentic AI workloads consuming inference capacity.

Framework's new flagship rewrites the portable PC playbook with 20-hour battery life, Ubuntu out of the box, and full cross-generational modularity.

Premium subscribers can now pin topic-specific feeds to their home screen, powered by Grok's semantic understanding of every post on the platform.

A rundown of every major announcement from OpenAI's livestream event, from new models to developer tools.

The new model arrives just seven weeks after GPT-5.4, offering improved coding, computer use, and scientific research capabilities.

Google unveils two distinct TPUs at Cloud Next 2026, with the 8t targeting frontier model training and 8i optimized for the agentic AI workloads consuming inference capacity.

Premium subscribers can now pin topic-specific feeds to their home screen, powered by Grok's semantic understanding of every post on the platform.

A rundown of every major announcement from OpenAI's livestream event, from new models to developer tools.

A hyper-realistic AI-generated desktop screenshot hints at what OpenAI plans to unveil at noon Pacific today.

A 15-million-parameter world model trained on a single GPU plans 48x faster than foundation-model competitors and encodes real physics in its latent space.