The Future of AI Isn't One-Size-Fits-All
summary
Claims that 70% of AI spend flows to four US hyperscalers and that decentralised vendors deliver provable compliance via cryptography are central, with concrete examples: companies pairing GPUs with CPUs to build hardware-level attestations (working with Intel, Nvidia, Dell and the Hideera ledger), a largest decentralized GPU service offering regional data residency and lower cost, and integrations with Snowflake, Databricks and Palantir. The surprising number, the hardware proof approach, and live integrations decide whether this is worth watching.
AI-written summary · about
chapters (13)
- 0:00Introduction
- 1:07Cloud feudalism and decentralized AI
- 3:56Web 2 vs. Web 3 in AI
- 6:24Proof of control in AI stacks
- 9:34Portfolio AI approach
- 12:45Redesigning internet for agentic AI
- 15:56Side-by-side decentralized AI
- 18:21Industry associations in AI adoption
- 21:07Developers shaping AI standards
- 24:15Europe’s AI compliance opportunity
- 26:45Healthy AI ecosystem in five years
- 28:29Developers’ demands for AI future
- 30:13Learn more about Advanced AI Society
related talks
28mThe Future of AI Won’t Be Centralized
44mGPU-less, Trust-less, Limit-less: Reimagining the Confidential AI Cloud
46mThe tech giants in AI - who is going to win the battle?
08mshares insights on AI and regulatory compliance
06mThe AI Compute Crisis & Web3 Opportunity
49mThe Role AI & Crypto Play in Consumer Facing Applications
10mWhy the Future of AI Depends on Blockchain
17mState of AI 2025 – Opening Keynote