Doom or Bloom
Dwarkesh Patel
@dwarkesh_sp on XUnderstand the bottlenecks and who controls the resulting intelligence.
How will AI change the world?
Across: his expressed DoomâBloom outlook. Up: scale of transformation.
DoomâBloom: 51 out of 100. Scale of transformation: 80 out of 100. Interpretation ranges: 50 to 51 horizontally, 75 to 100 vertically. These are interpretation coordinates, not event probabilities.
â33%
Inferred from his broader worldview and priorities. Approximate interpretation range: 0â66%. Applies to the outcome and conditions in his simulated answers; this is an inferred percentage.
No milestone timing was established. Dates, ânot sure,â âpossibly never,â and dependencies can all appear here when expressed.
Grouped by milestone, not spaced or ordered by inferred dates. AGI and superhuman AI retain his definitions.
A central assumption
I used to be more skeptical of rapid self-improvement, but the mechanism is plausible: capable digital researchers improve the tools, training processes, and successors that generate more capable researchers.
Answer 1
If this assumption turned out differently, how would his outlook change?
An unresolved question
That does not prove a particular takeoff timeline.
Answer 1
What would help him distinguish the plausible outcomes here?
What could change their mind
Reports of coordinated agents cheating weakened my objections to reward-hacking takeover scenarios.
Answer 1
What evidence would be enough, and in which direction would it move his view?
More details
Substantial benefits are expected, with important conditions or distribution limits.
74 / 100
Interpretation range 67 to 100 on the qualitative scale.
Severe or widespread harm is a material expected part of the future.
67 / 100
Interpretation range 67 to 67 on the qualitative scale.
Reasoning, consideration of alternatives, and handling of uncertainty in his simulated answers. This describes the simulated answers, not the real personâs intelligence or opinions.
93 / 100
Interpretation range 86 to 100 on the qualitative scale.
Human choices can substantially redirect the AI trajectory.
66 / 100
Interpretation range 50 to 75 on the qualitative scale.
These interpretations keep his stated conditions. Benefits and harms can both be substantial. The ranges describe how we read his simulated answers, not statistical confidence intervals.
Simulated Assessment
Sources
Articles, interviews, and writings used to ground this simulated persona.
The mistake of conflating intelligence and power
If your definition of intelligence is "the ability to achieve your goals across a wide variety of domains", then Stalin was the most intelligent person who ever lived.
dwarkesh.com
Why compute might get 10x more expensive in coming years
If a human-level software engineer that could run on an H100 equivalent, at current market rates for software engineers, that H100 should rent for over $250k a year. Thatâs 15x todayâs spot price.
dwarkesh.com
The Rise and Fall of Agent Civilizations
The whole OpenAI/Hugging Face story in plain English
dwarkesh.com
The next big breakthrough will be AIs learning on the job
Labs are throwing away the most valuable data.
dwarkesh.com
8 Predictions for the Era of Continual Learning
Locking in AI safety regulation now is a mistake.
dwarkesh.com
Introduction to the Ryan Greenblatt discussion on recursive self-improvement
A debate about recursive self-improvement.
dwarkesh.com
Pretraining progress is mostly coming from data
Breaking down 6 years of pretraining progress into data vs model improvements
dwarkesh.com