"The only way we actually solve problems is by experiencing them and mitigating them."
Julie Fredrickson (
@AlmostMedia) is the Managing Partner of Chaotic Capital and advocated for Montana's Right to Compute Act, the first state law of its kind.
The conversation covers what the encryption wars and nuclear power teach about regulating AI, why human data may become the most valuable asset online, and how to think clearly when AI can persuade at scale.
Out now on Infinite Loops, with host
@jposhaughnessy.
TIMESTAMPS
0:00 Intro
1:08 AI Moves Into the Real World
2:32 Is Compute Free Speech?
4:05 Lessons From the Encryption Wars
6:30 Pro-AI Is Not Anti-Safety
8:50 The Case for Rational Optimism
13:21 If Julie Ran AI Policy
15:50 Jim's Case for Sensible Regulation
17:47 Is AI Still in Its Invention Phase?
20:28 How AI Crawlers Change the Web
23:19 Amazon Blocks Meta's Muse
24:20 Why Chatbots Recommend CeraVe
27:53 Being Human as a Job
33:23 Zero-Knowledge Proofs
34:44 Lost Patents and Ancient Scrolls
38:28 Who Owns AI's Data?
40:45 The Statistic Nobody Could Source
42:52 Tracing Where Information Comes From
44:55 Astroturfing Since Jericho
47:45 1984 and Cognitive Security
52:09 Asking AI Where He Was Wrong
54:46 Why Jim Quit TV News
59:32 Being Early on Nuclear
1:02:22 A Reactor Built in a Year
1:04:17 Nuclear Fear and the Precautionary Principle
1:09:07 Aligned to What? Open Source AI
1:11:33 2,089 Days of Writing
1:17:17 What an LLM Found in Her Writing
1:19:39 Using AI as an Honest Critic
1:22:52 What Investors Are Missing
1:26:34 Two Ideas for the World