Grendel1209's Shortform

Grendel1209

This is a special post for quick takes by Grendel1209. Only they can create top-level comments. Comments here also appear on the Quick Takes page and All Posts page.

A few tips on public communication on x-risk now that the issue is salient and there will likely be increasing windows for action in the coming months and years as we continue to see demonstrated warning shots, job loss, and that the issue has gained public attention:

Communicate in ways your audience will understand. If you're speaking to laypeople about AI use simple analogies. For example people will not follow you if you say "we will get recursive self improvement to super intelligent AI that builds novel nanotech". Instead go with simple analogies "AI is like digital nuclear weapons combined with offshoring for jobs". People intuitively understand these things.
Frame your ask as middle of the road: People have a natural tendency to want to pick the "middle ground" approach, frame your policy this way. "The AI companies themselves say their products may result in human extinction, can you imagine if a terrorist group was building a super-virus that might cause human extinction? The U.S. would spend trillions to stop it. {Insert policy ask here} seems like the only reasonable thing to do in this circumstance.
Don't get sucked into providing specific scenarios: People will poke holes in specific stories to make themselves feel better. Making a simple case that the labs say we are one track to create superintelligence, that we aren't on track to control it, that this could have very bad outcomes and that many leading scientists agree is plenty.
Appeal to Authority and position as Scientists vs Industry: Yes this is a logical fallacy, yes this is also very effective for communicating with the public. The chances of explaining the incredible nuance here to a level someone will fully understand it who doesn't have a background is near zero. They are hearing opinions from Zuckerburg, Jensen Huang, Sam Altman, etc. Explain it as "We should trust the people not financially motivated, many leading scientists believe near term AI extinction risk is real including Dr. Max Tegmark at MIT, Dr. Geoffrey Irving who led the UK governments AI risk institution, Dr. Yoshua Bengio and Dr. Geoffrey Hinton who invented foundations underlying the technology.
Dress and Present Professionally: Wearing odd clothes, not seeming "normal" will reduce your credibility with the public. Humans have a natural heuristic to judge someone's credibility based on their appearance.

Communicate in ways your audience will understand. If you're speaking to laypeople about AI use simple analogies. For example people will not follow you if you say "we will get recursive self improvement to super intelligent AI that builds novel nanotech". Instead go with simple analogies "AI is like digital nuclear weapons combined with offshoring for jobs". People intuitively understand these things.
Frame your ask as middle of the road: People have a natural tendency to want to pick the "middle ground" approach, frame your policy this way. "The AI companies themselves say their products may result in human extinction, can you imagine if a terrorist group was building a super-virus that might cause human extinction? The U.S. would spend trillions to stop it. {Insert policy ask here} seems like the only reasonable thing to do in this circumstance.
Don't get sucked into providing specific scenarios: People will poke holes in specific stories to make themselves feel better. Making a simple case that the labs say we are one track to create superintelligence, that we aren't on track to control it, that this could have very bad outcomes and that many leading scientists agree is plenty.
Appeal to Authority and position as Scientists vs Industry: Yes this is a logical fallacy, yes this is also very effective for communicating with the public. The chances of explaining the incredible nuance here to a level someone will fully understand it who doesn't have a background is near zero. They are hearing opinions from Zuckerburg, Jensen Huang, Sam Altman, etc. Explain it as "We should trust the people not financially motivated, many leading scientists believe near term AI extinction risk is real including Dr. Max Tegmark at MIT, Dr. Geoffrey Irving who led the UK governments AI risk institution, Dr. Yoshua Bengio and Dr. Geoffrey Hinton who invented foundations underlying the technology.
Dress and Present Professionally: Wearing odd clothes, not seeming "normal" will reduce your credibility with the public. Humans have a natural heuristic to judge someone's credibility based on their appearance.