Hello,

Last week, we hosted a subscriber-only discussion on one of the biggest questions shaping AI right now: who gets access to the compute needed to build, train and scale the next generation of models.

As demand for AI infrastructure surges, data centers are delayed, GPUs remain scarce and startups are cutting deals with the companies that control capacity. Investors, operators and AI leaders are already rethinking what growth looks like when compute access becomes a competitive advantage.

In this conversation, The Information’s Anissa Gardizy sat down with Sachin Katti of OpenAI, Anjney Midha of AMP PBC and Jeremie Eliahou Ontiveros of SemiAnalysis to break down:

• What the compute crunch actually looks like on the ground
• Whether data center delays are as bad as rumored
• How hardware diversity is reshaping the market beyond Nvidia
• What startups can build to solve AI’s infrastructure bottlenecks

Subscribe $399 $299 and save 25% on the first year to watch the full conversation now.