Ai2's MolmoBot system trains robots entirely in virtual environments, outperforming physical-data models while releasing 230,000 scenes and 42 million annotations to the public.
Aderant's generative AI evaluation tool marks latest push to automate human judgment in professional services, raising questions about algorithmic management's expanding reach.
As hyperscalers commit $2 trillion to GPU infrastructure through 2030, emerging chip designs promise 100x efficiency gains—potentially rendering dense matrix multiplication obsolete.
Companies that embedded AI into workflows now confront a metering challenge: tracking employee consumption of tokens, the computational units that determine AI bills.
Federal lawmakers are preparing legislation to override state-level AI regulation, as Ohio and Tennessee lawmakers grapple with how to govern the technology without clear consensus.
Chipmaker Nvidia partners with startup Catches to launch RealFit, a platform using physics simulation to improve digital garment fitting, as luxury and retail sectors seek practical AI applications.
Research lab's new system automates creative brief to final execution, while compressed LLMs raise questions about consensus-driven output and creative homogenization.
After commanding 90% of the AI accelerator market, Nvidia faces a strategic shift as Google, Microsoft, and Meta develop cheaper alternatives, forcing the company to abandon its 'one chip fits all' model.
