8 Ways To A.I. Doom! With Jim Rutt
About this episode
Jim Rutt returns to walk through a comprehensive taxonomy of AI capability — weak, narrow, broad/general (AGI), and superintelligence (ASI) — using the Wozniak coffee-making test as the benchmark for true AGI and noting current self-driving cars and LLMs remain narrow/broad, not general. He traces the technical lineage (deep learning neural nets running on gaming GPUs, transformer architecture) that made today's generative AI boom possible, and predicts near-term job disruption will hit mid-tier white-collar work (law, accounting, customer service) before blue-collar trades. The conversation turns dark: paperclip-maximizer-style superintelligence risk, AI-enabled warfare (loitering munitions, drone wingmen), state surveillance dictatorships, hyper-personalized political propaganda, AI scams, and — Rutt's biggest near-term concern — "epistemic decay," where an internet flooded with AI-generated "sludge" erodes society's collective ability to separate fact from fiction faster than institutions can adapt. He argues US governance has been dysfunctional since roughly 1994 and is now dangerously overrun by accelerating AI-driven complexity, floating "liquid democracy" as a long-shot fix. The episode closes with a genuine on-air disagreement between Rutt and Vance over Bitcoin: Rutt calls it a "self-fueled bubble" and a worse store of value than cash in a mattress, while Vance defends it as a hedge against currency debasement and government "printing time."
“You mentioned how these things hallucinate... what it's actually doing is just predicting the next word in the sequence... The reason I'm pointing this out, when people give a hard date, be skeptical.”
“Bitcoin makes people at least think they might be able to beat that curve... this idea of storing the ability to invest is fallacious... it's a self fueled bubble... eventually people are gonna all dump Bitcoin and say this is absurd and ridiculous.”
“An economy consists of basically six things, production, distribution, consumption, savings, investment and innovation... put your money into productive assets... Bitcoin very, very bad.”
Key moments
- Rutt lays out the weak → narrow → broad/general → superintelligence AI taxonomy, using chess and Go game-playing programs as the canonical narrow-AI example, and introduces the Wozniak "make coffee in a random kitchen" test for true AGI (windows).
- Rutt cautions that AGI hard-date predictions (e.g., Kurzweil's 2029 singularity) should be treated skeptically given the wide range of credible expert estimates (windows).
- Explanation of the technical breakthrough — running deep learning on gaming GPUs plus Google's transformer architecture (BERT) — that made today's generative AI wave possible and explains why it happened "all at once."
- Rutt describes hiring an MFA short-story writer as his company's prompt engineer, illustrating a brand-new profession created by the AI wave and his prediction that liberal-arts skills would find new tech-industry value (craft).
- middle ~46-49%: Discussion of which jobs AI will disrupt first — mid-tier white-collar roles (junior lawyers doing discovery, entry-level accountants, customer service) rather than blue-collar trades, with a real example of a law firm cutting new associate hires from 10-12/year to 4.
- Rutt outlines "bad actors doing specific bad things with narrow AI" — China's AI-driven surveillance state, hyper-personalized political propaganda microtargeting individuals via LLMs and data brokers, and AI-enabled financial fraud/deepfake scams (mob).
- Rutt's "epistemic decay" thesis — the internet is flooding with AI-generated "sludge" (fake websites, regurgitated SEO content) that will increasingly train the next generation of LLMs on their own slop, degrading collective sense-making faster than institutions can respond (trust).
- Rutt's argument that US governance has been dysfunctional since ~1994 and is being overwhelmed by AI-accelerated complexity, proposing "liquid democracy" as a structural fix while rejecting the idea of handing decision-making authority to AGIs outright (institutions).
- The Bitcoin debate erupts — Rutt calls it "a self-fueled bubble" driven by collective hallucination about scarcity, arguing money should never be treated as a static store of value (bubbles).
Notable quotes
“So many faults, parts of that argument... this idea of storing the ability to invest is fallacious as it turns out... it's a self fueled bubble.”
“The next step is called artificial super intelligence, which is something that's a computer doing work that humans used to do that's now way better than any human has ever been at many of those tasks.”
“I believe that the human brain evolved to be able to handle a certain number of inbound attention requests per day... when we start getting overlayed with more and more and more calls on our attention, we are essentially becoming depleted.”
Bring this conversation to your organization. Vance Crowe speaks to conferences, boards, and leadership teams on escaping information bubbles and economic uncertainty.
Book Vance to speak