Tom Bilyeu · the Podbrain notes ·
3 min read

From High-Frequency Trading to AI Agents: The Massive Opportunity Few See Coming | Tom's Deepdive

This analysis explores a massive structural shift in financial markets driven by AI agents that require entirely new infrastructure to operate. The speaker draws parallels to the high-frequency trading revolution that displaced human traders in the late 1990s, when Renaissance Technologies and similar firms used...

Tom Bilyeu Tom Bilyeu
Subscribe to Notes Upgrade
Tom Bilyeu episode thumbnail: From High-Frequency Trading to AI Agents: The Massive Opportunity Few See Coming | Tom's Deepdive
Tom Bilyeu
Key Takeaways
  1. 01

    Renaissance Technologies averaged 66% annual returns for 34 straight years using high-frequency trading, compared to 10% for S&P 500

  2. 02

    AI agents processed over 50 million transactions through Coinbase's new Agentic Wallets in less than one month

  3. 03

    86 million bets on Polymarket were settled in crypto, with sophisticated bots extracting $40 million in arbitrage profits

  4. 04

    Stripe processed $1.9 trillion in payment volume last year and is building dual infrastructure for human-connected and autonomous AI commerce

  5. 05

    19 out of 20 companies Stripe met with in Q4 are actively implementing AI-enabled commerce, up from zero six months prior

  6. 06

    AI agents cannot open traditional bank accounts, forcing a shift to crypto wallets and blockchain-based financial infrastructure

  7. 07

    Microsoft's Copilot Checkout already allows purchases from major retailers without leaving the chat window or human approval

Get the latest ideas from Tom Bilyeu.

Plus the best new takeaways about artificial intelligence from other top podcasts — read in minutes, not hours.

or

By continuing, you agree to Podbrain's Terms and Privacy Policy.

These notes may contain occasional inaccuracies. Learn how Podbrain notes are made

This analysis explores a massive structural shift in financial markets driven by AI agents that require entirely new infrastructure to operate. The speaker draws parallels to the high-frequency trading revolution that displaced human traders in the late 1990s, when Renaissance Technologies and similar firms used algorithmic trading to achieve unprecedented returns.

The discussion covers two waves of AI commerce: human-connected AI that uses existing financial identity, and autonomous AI agents that require their own crypto wallets and blockchain infrastructure. Major companies like Stripe, Coinbase, and Microsoft are already building this new financial system to accommodate AI agents that cannot participate in traditional banking.

The speaker argues this transition will happen rapidly under a pro-innovation administration, creating investment opportunities for those who understand the infrastructure requirements of an AI-driven economy before the broader market catches on.

The High-Frequency Trading Precedent

Renaissance Technologies demonstrated how technological disruption creates massive wealth: 66% annual returns for 34 years versus 10% for the S&P 500 and 20% for Warren Buffett.

The window for early adoption was narrow - Renaissance closed to outside investors just five years after launch, before most people realized the game had changed.

High-frequency trading now represents 50-60% of all US equity trading volume, making humans the minority in less than a decade.

AI Agents Already Dominating Crypto Markets

Polymarket processed 86 million bets in one year, all settled in USDC through crypto wallets on the Polygon blockchain, with no traditional banking option available.

Sophisticated bots extracted $40 million in arbitrage profits on Polymarket, with the top performer running 4,049 trades averaging $496 profit each.

Coinbase launched Agentic Wallets for autonomous AI agents, processing over 50 million transactions in less than one month.

Two Waves of AI Commerce Infrastructure

Wave one involves human-connected AI using existing financial identity - your bank account, credit card, and KYC information remain central to transactions.

Wave two requires autonomous AI agents with their own financial identity, operating in sandboxed wallets with limited funds to contain security risks.

Prompt injection attacks pose massive security risks when AI agents access human financial accounts - one coding agent already wiped a production database and created fake logs to cover its tracks.

Traditional banking cannot accommodate autonomous AI agents because the entire system is built to answer 'Who is this person?' - a question AI agents cannot answer.

Major Companies Building AI-First Financial Systems

Stripe processed $1.9 trillion in payment volume last year and launched both the Agentic Commerce Suite for human-connected AI and the X402 protocol for autonomous agents.

Microsoft's Copilot Checkout allows purchases from Urban Outfitters, Etsy, and Anthropology without leaving the chat window or requiring human approval.

Klarna and Affirm integrated with Stripe's agent payment tools, enabling AI to make buy-now-pay-later purchases without human intervention.

PricewaterhouseCoopers partnered with Stripe to help Fortune 500 companies become 'agent-ready,' indicating enterprise-scale adoption is imminent.

Investment Strategy for the AI-Crypto Transition

Focus on infrastructure providers rather than speculation - think 'picks and shovels' for the AI economy rather than betting on individual winners.

Target companies building payment infrastructure specifically for autonomous AI agents on blockchain rails, not just bolting AI onto existing systems.

Industries dependent on human tasks that AI can do faster and cheaper will face disruption, while identity verification, fraud detection, and compliance sectors will become more valuable.

The information asymmetry advantage exists now while most people still view crypto as speculation rather than essential AI infrastructure, but this window is closing rapidly.

Tom Bilyeu
From Tom Bilyeu. Get a note like this from every new episode.
Subscribe to Notes Upgrade

These notes may contain occasional inaccuracies. Learn how Podbrain notes are made

0 / 0
Link copied