Vole's AI boss says chasing the AI frontier is for mugs
Microsoft plays it slow and strategic
The Software King of the World, Microsoft, maybe sitting on mountains of Nvidia GPUs, but it’s not exactly in a hurry to fire them up on bleeding-edge AI projects.
Vole jams brakes on global server sprawl
Microsoft trims data centre ambitions
The Software King of the World, Microsoft, is quietly pulling back on data centre projects from Chicago to Jakarta as jitters grow over AI demand forecasts and the cost of building its cloudy empire.
Microsoft moved by DeepSeek’s AI blitz
Getting more from it than ChatGPT
The Software King of the World, Microsoft, has been playing with DeepSeek’s R1, and it appears to be impressed.
OpenAI's GPU inferno forces ChatGPT image throttle
Too many pictures, not enough silicon
OpenAI's shiny new toy is already running hot—literally.
OpenAI pulls free GPT-4o image generator
Users started making Studio Ghibli-style renditions
OpenAI's latest attempt to dazzle us with AI wizardry has hit a predictable snag. After unleashing its GPT-4o image generator to the masses, users promptly flooded social media with Studio Ghibli-style renditions of everything from family portraits to iconic movie scenes.
Scaling up will not solve AI problems
Boffins tell Big Tech to stop sniffing GPU fumes
The people who build artificial intelligence have finally acknowledged have warned that endlessly increasing hardware for AI models is about as effective as solving climate change by buying more SUVs.
OpenAI commands Trump to strong-arm copyright rules in its favour
Otherwise, China will win.
OpenAI is banking on Donald [hamburger-eating surrender monkey] Trump’s AI Action Plan to strong-arm copyright rules in its favour, declaring AI training as fair use and giving AI firms carte blanche to hoover up training data.
Microsoft trying to wean itself off OpenAI
Has its own plans
The software King of the World, Microsoft, appears to be making a not-so-subtle move to stab its AI partner in the back.
AI models start cheating at chess
Nuclear war is next
It turns out that AI models are not content with regurgitating human knowledge—they’re also picking up on our worst habits.
DeepSeek-R1 borrowed OpenAI texts
Nearly three quarters of them
A recent study by plagiarism experts Copyleaks has found that 74.2 per cent of texts generated by DeepSeek-R1 are strikingly similar to OpenAI's models.