Crossposted from the EA Forum.
Videos like this one by Kurzgesagt show that it is not wholly impossible to present x-risks to a more general social media audience. It seems to me that content like this paired with a concrete call to action like contacting your representative or donating to some org could be very high impact.
There are also quite a few influencers with tech-savvy audiences and who already dislike AI. Recruting them to help with AI x-risk seems potentially very valuable.
Note that, although I’ve focused on AI x-risk, I think this is more generally applicable. If we want to convince people that x-risks are a problem, I think social media is probably the fastest and cheapest way to do that.
Yes, that video is excellent. It is also over 16 minutes long to give a very cursory explanation that x-risk is even a thing. It isn’t literally true that you only get five words, but it’s also true that most people won’t watch a video that long unless they’re already quite interested in the topic or the person.
I do agree there is a lot of potential for better, simpler explanations of x-risk, targeted at a wider variety of audiences, and that social media is likely an important part of how that should be distributed. However, that seems to me to be only about as much of a suggestion as “television” or “radio” would have been fifty or ninety years ago.
I find that the typical range of social media content is so frequently overwrought in its claims of massive (good or catastrophic) impacts of all kinds of things that people naturally discount what they hear to a huge degree. Talk of global extinction gets rounded down to ‘something kinda bad might happen somewhere eventually.’ I also think that many of the people who ‘dislike’ AI discuss it in a way that does not give me much confidence that they understand x-risk, or are willing to invest in developing ways of accurately conveying an understanding of x-risk.