Julie Fredrickson, Managing Partner of Chaotic Capital, thinks there is a good chance that verified human data becomes one of the most valuable assets online as AI agents read more of the web than people do. She also advocated for Montana's Right to Compute Act, the first state law protecting people's right to use computers and AI tools.
Three years after Episode 181, Julie and Jim O'Shaughnessy cover what nuclear power and the encryption wars teach about regulating AI, why chatbots keep recommending the same sunscreen, Amazon blocking Meta's Muse shopping agent, and why proving who is human online may soon matter for almost everything. On watching young nuclear companies move: "We can build websites fast is one thing. We can build nukes fast is a whole other ball game."
KEY INSIGHTS
- Human data may become the scarce asset. As bots write more of the internet, Julie thinks there is a good chance that verified human content, and people who tell stories well, grow more valuable.
- Chatbots repeat what forums say. Ask an AI for the best sunscreen and, Julie says, the answer is often CeraVe, because the brand dominates skincare threads on sites like Reddit.
- Compute is speech and property. Julie sees the math behind computing as a form of speech, and Montana's law makes the state show a compelling interest before restricting how people use computers.
- Safe and fast is probably more likely than safe and slow. Julie watched young nuclear companies take new reactors critical once a few regulatory hurdles came down.
- Write down where you were wrong. After 2,089 days of daily writing, Julie says the record shows how often she is wrong, and Jim has caught his own journals contradicting memories he would have sworn to in court.
RIGHT TO COMPUTE
Read Montana's Right to Compute Act: https://mca.legmt.gov/bills/mca/title_0020/chapter_0100/part_0020/sections_index.html
TIMESTAMPS
0:00 Intro
1:08 How AI Moved Into the Physical World
2:32 Is Compute Protected by the First Amendment?
4:05 What the Encryption Wars Teach About AI Rules
6:30 Pro-AI Does Not Mean Anti-Safety
7:37 Safe and Fast vs. Safe and Slow
8:50 Rational Optimism vs. AI Doom
13:21 If Julie Were the AI Czar
15:50 Jim Pushes Back: Wall Street Before Regulation
17:47 Are We Still in AI's Invention Phase?
20:28 AI Crawlers and the Machine-Readable Web
23:19 Amazon Blocks Meta's Muse Shopping Agent
24:20 Why AI Chatbots Keep Recommending CeraVe
27:53 Being Human as a Job Category
33:23 Zero-Knowledge Proofs and Proving Identity
34:44 Lost Soviet Patents and the Herculaneum Scrolls
38:28 Who Owns the Data Behind AI Answers
40:45 The Statistic Nobody Could Source
42:52 The Open Information Project and Data Provenance
44:55 Astroturfing, Snow Crash and Utilitarianism
47:45 The Panopticon, 1984 and Cognitive Security
50:04 Will Gen Alpha Handle Propaganda Better?
52:09 Asking AI Where He Was Wrong
54:46 Why Jim Stopped Watching TV News
59:32 Being Early on Nuclear and Valar Atomics
1:02:22 A Nuclear Reactor Built in a Year
1:04:17 Nuclear Fear, Data Centers and the Precautionary Principle
1:09:07 Aligned to What? Open Source AI and Linux
1:11:33 2,089 Days of Daily Writing
1:14:31 How Memory Rewrites the Past
1:17:17 What an LLM Revealed From Her Archive
1:19:39 Using AI as an Honest Critic
1:22:52 What Investors Are Missing: Digitization and Compute Pricing
1:26:34 Two Ideas for the World
ABOUT THE GUEST
Julie Fredrickson is the Managing Partner of Chaotic Capital, an early-stage venture fund. She previously founded Coutorture Media, acquired by Sugar Inc., and Stowaway Cosmetics. She advocated for Montana's Right to Compute Act and has written publicly every day since January 1, 2021.
ABOUT THE HOST
Jim O'Shaughnessy is the founder and CEO of O'Shaughnessy Ventures and the author of the New York Times bestseller What Works on Wall Street.
He founded O'Shaughnessy Asset Management and hosts Infinite Loops.
