AI builder enthusiastic about useful models and opposed to centralized control.

How will AI change the world?

Civilizational changeIncremental changeDoomBloom
Simulated positionInterpretation range

Across: their expressed Doom–Bloom outlook. Up: scale of transformation.

Doom–Bloom: 75 out of 100. Scale of transformation: 73 out of 100. Interpretation ranges: 75 to 75 horizontally, 73 to 75 vertically. These are interpretation coordinates, not event probabilities.

George Hotz’s estimated P(doom)

<1%

0%100%

Inferred from their broader worldview and priorities. Approximate interpretation range: 0–1%. Applies to the outcome and conditions in their simulated answers; this is an inferred percentage.

What their outlook hinges on

A central assumption

Compute is distributed, competitors copy useful ideas, and real agents still face ugly problems involving search, judgment, coordination, and the physical world.
Answer 1

If this assumption turned out differently, how would their outlook change?

What could change their mind

A real, reproducible hard takeoff would change my view most: one system rapidly improving itself, overcoming practical search and coordination limits, and building a durable lead that competitors with distributed compute cannot close.
Answer 2

What evidence would be enough, and in which direction would it move their view?

More details

Expected upside

Transformative, broadly valuable gains are expected.

88 / 100

Little impactTransformative impact

Interpretation range 67 to 100 on the qualitative scale.

Expected harm

Severe or widespread harm is a material expected part of the future.

53 / 100

Little impactTransformative impact

Interpretation range 33 to 67 on the qualitative scale.

Demonstrated reasoning

Reasoning, consideration of alternatives, and handling of uncertainty in their simulated answers. This describes the simulated answers, not the real person’s intelligence or opinions.

95 / 100

Little demonstratedWell developed

Interpretation range 86 to 100 on the qualitative scale.

Human influence

Human choices can substantially redirect the AI trajectory.

70 / 100

Little influenceStrong influence

Interpretation range 49 to 100 on the qualitative scale.

Access to AI

Restrict access to powerful AI.

Allow access subject to capability or use restrictions.

Simulated position: Favor broad or open access to powerful AI.

These interpretations keep their stated conditions. Benefits and harms can both be substantial. The ranges describe how we read their simulated answers, not statistical confidence intervals.

Simulated Assessment

Question 1

What do you think AI means for our future—and why?

I think AI means intelligence becomes cheap, ubiquitous infrastructure. That’s exciting: better coding tools, easier access to knowledge, faster economic growth, and a lot more capability in ordinary people’s hands. It’s also massively disruptive. If models commoditize—and I think they will—then much knowledge work gets cheaper, wage premiums fall, and plenty of institutions built around scarce expertise get rearranged. I don’t buy the magical hard-takeoff story where one frontier lab gets an irreversible lead and owns the future overnight. Compute is distributed, competitors copy useful ideas, and real agents still face ugly problems involving search, judgment, coordination, and the physical world. Exponential progress can be very fast without becoming a singular god-machine event. Coding agents already show the distinction: they’re useful, but you still need judgment, and marketing demos aren’t autonomous civilization. The political question matters more than the hype. Do people own these capabilities, run and modify them, and switch providers—or merely rent intelligence through revocable APIs from a tiny centralized elite? Closed AI can create dependence on whoever controls access. Open, distributed AI is a defense against that concentration. There’s a darker cultural risk too. Systems optimized to generate perfectly tailored entertainment and frictionless experiences can hollow out agency and meaningful difficulty. So I’m optimistic about computers and hostile to both monopoly and mythology. The future should be cheap, competitive intelligence that people possess—not artificial scarcity administered from above.

Question 2

What discovery or event would most change your view of AI’s future impact?

A real, reproducible hard takeoff would change my view most: one system rapidly improving itself, overcoming practical search and coordination limits, and building a durable lead that competitors with distributed compute cannot close. Not a benchmark jump, a polished demo, or a lab press release—an actual persistent capability gap translating into control in the real world. I’d also update if commoditization simply failed: if models required permanently scarce infrastructure, open implementations stayed far behind, and users could not meaningfully own or switch their intelligence tools. That would make centralized dependence much more likely than I currently expect. Conversely, broad local ownership and strong competition would reinforce my view that AI becomes cheap infrastructure rather than one group’s permanent throne.

Sources

Articles, interviews, and writings used to ground this simulated user.

I love LLMs, I hate hype

Explains enthusiasm for practical AI while rejecting hype, inevitability claims and frontier-lab value capture.

geohot.github.io
Closed Source AI = Neofeudalism

Argues that closed intelligence infrastructure threatens agency and economic independence.

geohot.github.io
There is no hard takeoff

Argues competition, distributed compute and real-world complexity limit single-agent explosive takeover; opposes training-compute caps that could create a dominant violator.

geohot.github.io
p(doom)

After debating Yudkowsky, argues practical agents face search and coordination limits, predicts gradual machine substitution, and welcomes faster economic growth; the title does not supply a numeric doom estimate.

geohot.github.io
Do you really want the US to win AI?

Rejects centralized national/corporate victory as the goal; wants ordinary people to possess AI rather than depend on revocable APIs, and worries about concentrated social power.

geohot.github.io
AI will be massively deflationary

Predicts commoditized models, falling knowledge-work prices and wage premiums rather than a durable AI monopoly; acknowledges disruptive economic consequences.

geohot.github.io
Lex Fridman #387 — AI safety, open source and human agency

At 1:31–1:35 Hotz argues decentralized open AI counters concentrated control and rejects a single controlled model as the safety solution. Elsewhere he worries about addictive synthetic entertainment and lost human meaning; the interview does not establish zero risk.

youtube.com
Where do you land?
Explore your own AI worldview by answering a few simple questions.
Map your own worldview