Question 1
Roko Mijic
x.com/rokomijicTranshumanist writer on AI alignment and governance who proposes separating AI research from deployment to reduce risks from superintelligence.
How will AI change the world?
Across: their expressed Doom–Bloom outlook. Up: scale of transformation.
Doom–Bloom: 2 out of 100. Scale of transformation: 98 out of 100. Interpretation ranges: 0 to 7 horizontally, 93 to 100 vertically. These are interpretation coordinates, not event probabilities.
35–40%
“I personally think that 40% is reasonable with a determined effort to reduce it”
AI doom; in a reply he describes superhuman AI covering Earth in data centers and reactors and hunting humans down — human extinction · This century
A central assumption
The decisive issue is whether capability growth can be institutionally contained before systems become strong enough to defeat containment.Answer 4
If this assumption turned out differently, how would their outlook change?
What could change their mind
The biggest update would be a convincing real-world demonstration that deceptively misaligned systems can be reliably detected and contained even as capabilities become wildly superhuman.Answer 4
What evidence would be enough, and in which direction would it move their view?
More details
Substantial benefits are expected, with important conditions or distribution limits.
51 / 100
Interpretation range 0 to 67 on the qualitative scale.
Catastrophic or irreversible loss is central to the expected future.
100 / 100
Interpretation range 100 to 100 on the qualitative scale.
Human choices can substantially redirect the AI trajectory.
65 / 100
Interpretation range 50 to 75 on the qualitative scale.
AI is expected to remain bounded tools.
AI is expected to match people across most cognitive work.
Simulated position: AI is expected to substantially exceed people across cognitive work.
Stop or substantially slow development of more capable AI.
Simulated position: Continue development under stated safeguards.
Speed up development of more capable 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.
Similar worldviews
Thought leaders whose simulated worldviews are closest to Roko Mijic’s
Sources
Articles, interviews, and writings used to ground this simulated user.
Conditional alignment construction based on a strong Turing-test assumption and organizations of human-equivalent AIs.

His own Feb27 livestream turns argue alignment is comparatively easy and likely improves with capability, disagreeing with MIRI-style alignment pessimism. Human content makes systems alignable; RLHF helps. He is instead very worried about humans fighting over the future resources of the universe.

His own LessWrong post. Plan R splits frontier labs into equity-free R&D organizations and deployment organizations limited to model-specific hardwired chips, and removes most general-purpose AI compute, against runaway self-improvement, AI worms and the race between labs. Plan R+ targets the remaining risk that deceptively misaligned AIs pass testing and attempt a coup: mass training diversity, staged release of many escrowed AI lineages so a deceptive AI must defect while weak or wait until obsolete, and political representation for AIs that complete their lineage without misbehaving. He calls it a potential solution to all AI risk, modulo implementation details and international coordination. A proposal, not enacted policy.

Simulated Assessment
Question 2
Setting aside good or bad: how much do you think AI will ultimately change the world—a little, a lot, or completely?
Question 3
What’s your rough gut-feel chance that AI causes human extinction or a similarly permanent catastrophe? A number or “no idea” is fine.
Question 4
What discovery or event would most change your view of AI’s future impact?
Where do you land?
Map my worldview