Coinbase Cuts AI Spend by 50% | Kalshi's $40B Valuation & Impending IPO | The Year for SaaS Roll-Ups

Companies are aggressively scrutinizing AI expenditures as the focus shifts from "token maxing" to proving tangible revenue growth. The current market environment is prioritizing proven ROI over experimental spending, with a growing trend of companies pivoting to cheaper open-source models or consolidating underperforming software assets to extract value.

• Coinbase AI cut — the company reduced its AI spending by 50% by switching to open-source alternatives, proving that even major tech firms are prioritizing cost discipline over frontier model reliance [01:29].
• Proving AI ROI — leadership across the industry is struggling to demonstrate that increased spending on AI agents is leading to actual revenue lift or productivity gains [08:51].
• Model distillation — Anthropic is lobbying the government to address Chinese models training on their data, effectively trying to position the issue as a national security concern rather than just a contractual dispute [18:28].
• Microsoft performance — the company faces pressure due to decelerating Azure growth and the lack of a standalone consumer AI product to rival its competitors [33:53].
• Prediction markets — companies like Kalshi are pursuing massive valuations by targeting mass-market gambling—specifically sports and financial betting—rather than niche political event betting [40:06].
• Bending Spoons strategy — the IPO focuses on a "roll-up" model, acquiring tired, older software products and aggressively optimizing them to improve profit margins [43:22].
• VC-turned-CEOs — there is skepticism regarding billionaire investors taking on CEO roles, with doubts about whether they can maintain the necessary intensity required for a successful startup grind [1:00:02].

• How does the "distillation" of foundation models work according to the speakers?
• What are the two primary growth paths the speakers identify for prediction market platforms like Kalshi?