U.TodayU.Today

Crypto vs AI: Balaji Srinivasan's Counterintuitive Take

Lesedauer 2 min

While normally, artificial intelligence is expected to cause massive layoffs, it also can create a big demand for previously unimagined professionals. Cryptocurrency and blockchain are inevitable in handling dangerous imbalances associated with AI, entrepreneur, investor and The Network State author Balaji Srinivasan says.

"Crypto makes it real again": Silicon Valley legend Balaji Srinivasan

Artificial intelligence (AI) will create a massive number of jobs in proctoring and verification because it is extremely good at faking things. In turn, crypto, with its strict deterministic patterns, can be helpful in mitigating the dangers here, former Coinbase CTO Balaji Srinivasan says in a new episode of the a16z podcast.

a16z
@a16z

Balaji Srinivasan on the coming verification gap in an AI world.

“AI is going to create massive numbers of jobs in proctoring and verification because it’s so good at faking things.”

“AI makes everything fake, and crypto makes it real again.”@balajispic.twitter.com/65sKsnsCuv

Nov 28, 2025

With the evolving opportunities of large language models (LLMs) and the opportunities of AI tools to create fakes — in terms of media, data, science, from routine facts to sophisticated theories — blockchain tech can remain a beacon of verification that cannot be tampered with.

Even the most powerful AI cannot create a private key that would be relevant to a real Bitcoin BTCUSD address or an on-chain NFT that was not actually minted.

As such, blockchain-based data actually remains the only sort of data AI cannot fake or alter thanks to its technical flaws or "hallucinations":

AI makes everything fake, and crypto makes it real again

To prove the theory, Srinivasan recalled the FTX hack. If you ask Perplexity, a popular content-centric LLM, to provide you with a summary of the 2022 FTX collapse, on-chain transactions is the only verifiable data it relies on.

AI crypto capitalization collapsed in 2025

Diving deep into the context of the FTX hack, the reader can see exactly which transactions happened and when, and this data cannot be altered or removed.

As covered by U.Today previously, a number of applications at the intersection between AI and crypto launched in 2025 with the growing popularity of the Model-Context Protocol concept.

X402, an AI agent payments protocol by Coinbase and Google, showcased the opportunity to completely remove human interactions from the spending process.

Meanwhile, the euphoria around AI crypto tokens seems to have ended. In 2025 alone, the capitalization of this segment dropped by 74%, shrinking from $13.5 billion to $3.5 billion, CoinGecko's AI Agents crypto tracker says.