🕶️ Runway outfilms Google. Murati reboots

Runway, founded by NYU film-school grads, not Stanford CS royalty, is betting $5.3 billion that the path to general intelligence runs through video, not chatbots, and it's already shipping world models while Google watches.

The thesis is that if you teach AI to simulate physics from pixels, you unlock everything from robotics to drug discovery; the risk is that no one has proven video understanding actually scales to reasoning, and the compute bill makes OpenAI's Sora meltdown look like a rounding error.

Elsewhere, over $800M is landing in emerging sectors where the product isn't the model but the infrastructure around it, Turso killed its bug bounty after realizing 90% of submissions were AI slop, and Revolut handed 10,000 employees £1K each to hunt business accounts before a $200B IPO.

This Saturday: where capital is compounding, which bets on intelligence are actually differentiated, and what happens when the outsiders start outrunning the incumbents.

🕶️

Connect your email, calls, product data and more, and Attio instantly builds your CRM with enriched data and complete context. Whether you’re running product-led growth or enterprise sales, Attio adapts to your unique GTM motion.

Then Ask Attio to plan your next move.

Run deep web research on prospects. Update your pipeline as you work. Find customers and draft outreach emails. Powered by Universal Context, Attio's intelligence layer, Attio searches, updates, and creates across your data to accelerate your workflow.

Ask more from your CRM.

Ask Attio

Apparently, the path to challenging Google in AI starts with NYU art school grads, not Stanford CS legends. Runway’s art‑school outsiders scale from filmmaker tools to a $5.3B bet that intelligence lives in video, not text.

Their thesis: video-generation models are the shortest route to physics-aware world models. Runway has already ships first world‑model and is testing robotics and scientific simulations, treating “digital twins of the universe” as core infrastructure for everything from Hollywood pipelines to drug discovery.

The catch is compute and proof. No one has shown that video understanding scales into general reasoning, and training these models requires cluster-level access that even OpenAI struggled to justify, as its compute burn cautionary with Sora showed.

Runway’s real experiment isn’t just technical: it’s whether a revenue-disciplined, non–Bay Area culture can outrun incumbents in a race that looks, from the outside, purely about capital.