
Rakesh Gidwani
@rakeshgidwani
Co-Founder & CTO @ aion.xyz, Venture Partner & Advisor @ Protagonist.co
ID: 58987527
https://rakeshgidwani.com/ 22-07-2009 01:15:50
151 Tweet
244 Followers
375 Following

.McKinsey & Company & Company estimates that by 2030, data centers will need $7 trillion to scale AI… Two words: 👏 decentralize 👏 compute. We could build 100,000 hyperscale warehouses, or maybe we just launch a permissionless network at a fraction of the cost? Better yet, let users

"Analytics always expands to fill the available budget." Tristan Handy breaks this down with Jennifer Li and Matt Bornstein on the AI + a16z podcast: Save companies money on compute and they immediately find new workloads. This is Jevons Paradox and it explains the modern data



Excited to see where this goes Pluralis Research 🤝 We should chat!



.Eric Schmidt says AI could one day consume 99% of the world’s electricity. I think he’s right. Compute needs two things: electricity and water. And we’re scaling both faster than we’re scaling generation. If demand keeps rising, short-term? Electricity prices go up. Long-term?
