Balaji Srinivasan: Prove Correct, Not Just Go Direct
Summary
Balaji Srinivasan, angel investor and author, discusses how AI is transforming media and eroding trust, necessitating a shift from "go direct" to "prove correct." He highlights that as content creation costs approach zero, verification costs rise, leading to a breakdown in systems like journalism and hiring due to synthetic content. Srinivasan advocates for a new information stack built on cryptography, on-chain data, and verifiable records to establish provable truth, citing examples like Vitalik Buterin's proof of life and blockchain evidence in Chinese courts. He also critiques traditional media's historical biases and economic disruption by technology, proposing a future of decentralized, open-source citizen journalism and "on-chain media" to combat misinformation.
Key takeaway
For technologists and entrepreneurs building new media platforms or trust systems, you must prioritize verifiable, cryptographically-backed information over mere direct publishing. Focus on developing open-source, decentralized, on-chain media solutions to provide provable facts, rather than relying on traditional institutional assertions of truth. This approach is essential to counter AI-driven misinformation and rebuild public trust in an increasingly synthetic information landscape.
Key insights
Cryptographic verification and on-chain data are crucial for establishing provable truth in an era of pervasive synthetic content.
Principles
- "Prove correct" is superior to "go direct" for establishing truth.
- Easy to verify, difficult to fake is critical for systems dealing with strangers.
- AI is polytheistic and constrained, not a monotheistic AGI.
Method
Build systems on cryptography, on-chain data, and verifiable records to make truth provable. Utilize on-chain data feeds with AI for automated, unbiased journalism.
In practice
- Cryptographically sign letters of reference or endorsements.
- Use on-chain records (e.g., Etherscan links) to verify financial events.
- Implement "internet intermediate" models where users opt-in to community constraints.
Topics
- AI and Media
- Cryptographic Verification
- On-chain Data
- Decentralized Journalism
- Information Integrity
- Trust Systems
Best for: CTO, VP of Engineering/Data, Director of AI/ML, Entrepreneur, Investor, AI Ethicist
Related on AIssential
Editorial summary, takeaway, and curation by AIssential. Original article published by The a16z Show.