PrismML Launches 1-bit Bonsai 8B Model That Is 14x Smaller, 8x Faster Than Competitors
Even as models keep getting larger, some companies are moving models in the opposite direction — with some impressive results. Caltech-originated AI lab…
Pre-training Of Future Models Will Involve Models Taking Actions, Choosing Which Data To See Next: Google’s Jeff Dean
Pretraining AI models is thought by some to be fairly settled with set procedures and datasets, but it could change dramatically in the…
The Hidden Mistakes Startups Make Before They Even Launch
When people talk about startups, the conversation usually begins with product-market fit, funding, or growth hacks. But long before any of that matters,…
Claude Code Leak Was ‘Human Error’, No One Was Fired: Claude Code Creator Boris Cherny
Claude Code has been writing nearly 100% of its own code, but the leak of its codebase yesterday didn’t have anything to do…
Apart From Chess, Google DeepMind CEO Demis Hassabis Has Won Championships In Diplomacy, Poker & Pentamind
Google DeepMind CEO Demis Hassabis has won the Nobel Prize and is currently helping usher the world through the AI revolution, but he’s…
Used To Write Code Everyday, But AI Has Written All My Code Since December: Reddit CEO Steve Huffman
More and more CEOs are talking about how they’re back to coding — and AI is doing all of it for them. Reddit…
How WhatsApp Didn’t Have Code Reviews Until It Was Acquired By Facebook
WhatsApp was, by almost any measure, the most efficiently run engineering organisation in tech history. Jean Lee, engineer number 19 at WhatsApp and…
Facebook Founder Mark Zuckerberg Had Not Talked At All During Their First Meeting, Reveals Marc Andreessen
Facebook CEO Mark Zuckerberg had been known to be a reluctant public speaker in his early years — he’d once turned red and…
User Discovers That GitHub Copilot Edited An Ad Into Their PR, Company Responds
A routine typo fix turned into a corporate trust crisis for GitHub this week, after a developer discovered that GitHub Copilot had quietly…