“Superintelligence.com” Founder’s Plan To Get P(Doom) Below 0.01% — Debate with Dr. Craig Kaplan

In an 85-minute episode, Liron Shapira debates Craig Kaplan, who owns superintelligence.com and puts the chance of AI-caused catastrophe at about 10%. Kaplan describes a three-phase plan, built on networks of AIs that check each other, that Kaplan says could cut that risk a thousandfold.

Why it matters

The same thing, for different readers.

What it means for you

Craig Kaplan, who takes AI extinction risk seriously, argues millions of personal AIs checking one another could be safer than a few giant systems; no mechanism exists yet.

For researchers

Kaplan's safety argument rests on networked AIs monitoring each other faster than people can; whether they would do so, and share human values, remains open.

The question it raises

Kaplan's 0.01% target is the rate at which humans kill each other, which raises whether an AI that matches human violence counts as aligned.

In their own words

Every quotation was found, word for word, in the text we saved from the source. That shows the words are there; it does not make them true, or prove the speaker meant them as quoted.

From the publisher’s own transcript.

“I actually believe that you can build safe designs and we need to do that as quickly as possible. That's the accelerating part.”
Craig Kaplan, Founder, Superintelligence.comKaplan on where Kaplan differs from some in the doomer community, while explaining the plan and a P(doom) of about 10%.
Found word for word in the source
“Now my estimate has dropped. I'm sitting around 10%. And the reason for that is so many other people are talking about it.”
Craig Kaplan, Founder, Superintelligence.comKaplan gives a current probability of AI doom and says it fell from about 20% in 2020 as awareness grew.
Found word for word in the source
“when you have millions of entities working together in a network, they can check each other. And in fact, you can design the system so that they can track each other much faster than humans can think.”
Craig Kaplan, Founder, Superintelligence.comKaplan explains why a collective-intelligence network of AIs would be safer than one monolithic frontier model.
Found word for word in the sourcestarts inside a longer claim, right after “that”
“All of what we've just said, there's no mechanism yet. That's simply an aspirational goal.”
Craig Kaplan, Founder, Superintelligence.comKaplan on the 0.01% target, anchored to the human rate of killing, before turning to the proposed design.
Found word for word in the source

The record

Questions
Can AI systems act beyond what people intended or allowed? · Are AI companies putting stronger safeguards into practice?
People
Craig Kaplan (Speaks, founder, IQ Company; owner of superintelligence.com) · Liron Shapira (Speaks)
Also covered by
Doom Debates (YouTube)
What we read
We could read only part of the source.
Found by
The Radar, run of November 18, 2026, via pod-doom-debates (tier 1: the source’s own channel or an official record)
Cite this
Doom Debates, ““Superintelligence.com” Founder’s Plan To Get P(Doom) Below 0.01% — Debate with Dr. Craig Kaplan”, November 17, 2026. Entry in the Ctrl AI Base, https://ctrlai.com/base/doomdebates-2026-11-17-kaplan-pdoom-plan
Our notesRead the first 16 minutes of the publisher's transcript (the plan, the P(doom) figures, the network argument); the later debate on orthogonality, instrumental convergence and 'stoner' vs 'Caesar' AIs was not read, so the summary covers the opening only. The full VTT is in sources/f3163d13c961.txt; its cue timestamps broke the checker, so sources/f3163d13c961.plain.txt holds the quoted passages with timestamps removed and words unchanged. The transcript has no speaker labels; speakers are identified from the introductions and the answers.