Episode Details

Motivation: If we want to move from Plan D to Plan A or S, I believe the first step is to collectively agree on the problem. We are far from it, and there is a lot we can do.

Abstract:

1. We already know enough to act. I wish we were in a world where research was the bottleneck, but the main constraint on AI safety is no longer a shortage of clever policy ideas: best practices already exist and are not being applied or enforced, and a serious international (or even just national) regulatory regime would probably cut most of the risk.
2. They are not applied because awareness is low. The people who narrate and enforce AI policy mostly do not believe the problem exists. I estimate that a majority of the top ~100–1,000 most influential policymakers worldwide have never had a single serious conversation about catastrophic risk, and this is the main reason they are not worried[1]. Even among the civil-society organizations that showed up to the UN Global Dialogue, exactly one of the 1,534 written submissions mentions "takeover", and less than 1% mention x-risks explicitly.[2]
3. They've never had the conversation because our field under-invests in having it. Status rewards [...]

---

Outline:

(03:53) 1. -- The bottleneck is political will, not research

(04:08) What do I call "political will"?

(05:34) The best practices we already have are not being applied

(07:53) We need to go from plan D to plan A: more seriousness and coordination

(09:47) Winning requires a large fraction of the top ~100-1,000 to understand the problem, and we're far from that.

(12:04) 2. -- Almost nobody realizes how bad the situation is - hence no political will

(12:21) What the people in charge don't believe

(14:13) Among the NGOs, almost no one talks about x-risks

(15:26) The public is still apathetic, and when prompted, mis-prioritizes

(16:36) 3. -- Why is awareness so low?

(18:04) We're not in enough rooms

(18:37) If we won't name the risks, who will?

(21:51) Status rewards research; the work that moves policy goes unrewarded

(23:31) One exposure is not enough; repetition is how you convince

(25:39) The field is comically small

(27:38) The main objection: Let's just wait for a warning shot? I don't think this works.

(31:53) Other potential objections

(35:18) 4 -- What to do

(35:28) A. Do the direct work

(35:32) The bottom line: talk to more people!

(36:38) Contribute to open consultations

(37:02) Advocacy aimed at AI lab employees

(37:36) Explain what you truly believe instead of just making a brittle recommendation

(38:16) B. Build the credibility infrastructure

(38:20) Prepare to own the next crisis

(39:01) Pre-register what you expect to see

(39:38) Track the skeptics' predictions

(40:18) C. Communication at scale

(40:22) Viral communication

(40:54) Break the "sci-fi" stigma

(41:23) Invest in coordination mechanisms between AI safety advocates

(42:56) D. An opinionated list of research directions that I find most useful

(46:19) Addendum

(46:27) Political will is (or will be) the bottleneck for nearly every cause, not just AI safety

(47:36) We can automate the research, but not the consensus

The original text contained 40 footnotes which were omitted from this narration.

---

First published:
July 19th, 2026

Source:
https://forum.effectivealtruism.org/posts/nmQpZyybjHrunDxFT/the-current-bottleneck-is-p