Andrew Jones

Andrew Jones

x.com/dremnik

Founder, designer and engineer who argues that when AI makes execution cheap, the bottleneck shifts to clarity, judgment and design.

How will AI change the world?

Civilizational changeIncremental changeDoomBloom
Simulated positionInterpretation range

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

Doom–Bloom: 47 out of 100. Scale of transformation: 69 out of 100. Interpretation ranges: 42 to 52 horizontally, 49 to 76 vertically. These are interpretation coordinates, not event probabilities.

Andrew Jones’s P(doom) · inferred

≈10%

0%100%

Inferred from their simulated answers, not a number they gave. Plausible range: 2–31%.

How we estimate P(doom)

What their outlook hinges on

A central assumption

Producing code, analysis, drafts, and prototypes gets faster; deciding what should exist, which problem matters, what tradeoffs are acceptable, and whether the result is actually good remains stubbornly human work.
Answer 1

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

An unresolved question

What remains uncertain is whether that change produces more capable individuals and small teams, or more dependence on a few centralized systems.
Answer 3

What would help them distinguish the plausible outcomes here?

More details

Expected upside

Substantial benefits are expected, with important conditions or distribution limits.

65 / 100

Little impactTransformative impact

Interpretation range 67 to 67 on the qualitative scale.

Expected harm

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

58 / 100

Little impactTransformative impact

Interpretation range 33 to 67 on the qualitative scale.

Human influence

Human choices have meaningful but substantially constrained influence.

55 / 100

Little influenceStrong influence

Interpretation range 27 to 98 on the qualitative scale.

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.

Where do you land vs Andrew Jones?
Map your own AI worldview in about 3 minutes, then compare

Similar worldviews

Thought leaders whose simulated worldviews are closest to Andrew Jones’s

Sources

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

Simulated Assessment

Question 1

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

I think AI makes execution abundant while making clarity more valuable. Producing code, analysis, drafts, and prototypes gets faster; deciding what should exist, which problem matters, what tradeoffs are acceptable, and whether the result is actually good remains stubbornly human work. Better execution does not automatically generate better ideas. That shifts organizational advantage. When work depends on many serial judgments, adding more people can add coordination faster than it adds capability. Small, high-trust teams of broad generalists may therefore outperform larger organizations: they can hold a coherent product model, make decisions quickly, and use AI to cover more of the execution surface. But this is not a universal law; some futures could still be highly centralized, especially where capital, data, or distribution dominate. The larger uncertainty is human agency. “Past technologies created new jobs” is not a concrete account of what people will do if machines perform much of today’s knowledge work. I want descriptions of actual future roles, incentives, and institutions—not “AGI” used as a fog around the question. The interface matters too. Good collaboration should exchange coherent units of thought rather than expose every intermediate edit or imitate human presence too closely. I prefer systems that feel honestly machine-like and help people form, inspect, and act on clear intentions. The future turns less on raw generation than on whether humans retain the ability to choose worthwhile directions.

Question 2

Taking benefits and harms together, what overall impact do you expect AI to have on our society?

I don’t think the evidence supports a clean positive or negative forecast. AI could give small teams extraordinary leverage, make useful software cheaper, and let people turn clear intentions into working systems much faster. But it could also concentrate power around whoever controls capital, distribution, and the strongest systems, while making many existing forms of knowledge work economically uncertain. My main concern is the gap between execution and agency. If machines can do more of the execution, humans do not automatically inherit more freedom. Someone still chooses the problems, sets the incentives, and captures the value. A decentralized future of capable individuals and small teams is plausible; so is a centralized future where a few organizations mediate most consequential work. That is why vague claims about productivity or inevitable new jobs are inadequate. The overall impact depends on concrete institutions and interfaces: what work people actually do, who can direct these systems, and whether humans remain able to form and pursue worthwhile goals. AI increases capability, but capability alone does not determine whether society becomes better.

Question 3

How much do you think AI will ultimately change the world?

Enormously, but not in a way I can honestly compress into a percentage or a single “AGI changes everything” claim. If software execution becomes cheap and fast, then many assumptions behind organizations, professions, and product development change. A small group with strong judgment may be able to build what previously required a large company. Existing knowledge work could be reorganized around directing, evaluating, and integrating machine output rather than producing every intermediate artifact by hand. But technical capability does not specify the social outcome. The same leverage could broaden human agency or concentrate it among organizations controlling models, capital, and distribution. Nor is “new jobs will appear” an adequate answer. We need concrete accounts of what people will actually do, why anyone will pay for it, and where human judgment remains decisive. So I expect the change to be deep, especially in software-mediated work. What remains uncertain is whether that change produces more capable individuals and small teams, or more dependence on a few centralized systems.

Question 4

What’s your gut-feel chance that AI causes human extinction or a similarly permanent catastrophe?

I don’t have a defensible percentage. My concern is less a specific extinction forecast than the possibility that rapidly increasing capability outruns our ability to choose goals, design institutions, and preserve human agency. A permanent catastrophe could also look like durable concentration of control or dependence on systems people can no longer meaningfully direct—not only literal extinction. But I would not turn broad uncertainty into a precise probability. Before accepting a number, I would want a concrete causal account: which systems acquire which capabilities, who deploys them, what decisions become irreversible, and why existing actors cannot intervene. “AGI” by itself is not that account.
Where do you land?
Explore your own AI worldview by answering a few simple questions.
Map your own worldview

Where do you land?

Map my worldview