I'm a writer, not a technical person -- what I'm interested in trying to do is signal boosting ideas that within the community to the sort of general tech audience that reads hard sci fi novels, in the hopes of boosting serious interest and awareness around the subject, rather than painting a particular approach as the right approach.
I think that was a great comment :)
As for how this idea can be used -- I'd say that as a sort of artistic thing, as described it feels a little deus ex machina, which isn't necessarily a bad thing, its just I'm right now personally trying to come up with stories where by the time the AI is actually on the verge of being developed, enough right choices were made earlier that it is inevitable things go well, with the idea that what is valuable now is encouraging people to build the institutions and safety procedures into their system so that it doesn't come close. On the other hand that doesn't optimize for strong conflicts and climax, and I think your plan could do that really well.
We're both still just sort of guessing at what will actually help -- but signal boosting existing organizations like MIRI and CHAI and the idea of explicitly taking safety really seriously sounds promising to me.
One thing I do do in my Pride and Prejudice Variations is always write an afterward talking about how I wrote the book, and then ending with telling people that they should donate to Doctors Without Borders, something like that, explicitly having a simple call to action at the end of the novel probably is a good idea.
Who specifically do you think should act differently, and in what concrete way because they are more aware of the Beyond the Reach of God narrative?
I feel like there is a lot of dystopian literature out there, but relatively little about telling a story where there is a plausible path to escaping things going horribly wrong that then works. So I'm right now intentionally trying to come up with stories that sell an utopian path while signal boosting ideas that are being put forward in FHI papers and other parts of the community as ways to get there. For example the project I'm right now the most excited about has the working title of The Windfall Clause. Also the sci fi project that I already have written that is in this context is exploring ideas about the repugnant conclusion in a far future hard sci fi setting which is organized like Scott Alexander's archipelago, and where we managed to both get AI that did what we wanted, and then where we collectively didn't use it to murder ourselves. (Link if anyone is interested)
I do welcome ideas about stories that people think it would be a good idea if someone wrote. Though if it is about something going horribly wrong, I'd probably try to find a way to write a story where that nearly happens, but we find a smart way to avoid it happening.
Also, honestly, I think that all of the countries would reinvest as much as they need to maintain a strategic balance, and that is the actual problem requiring coordination.
Oh that's cool. I had known about Herzl being a central figure in Zionism, but not that he'd written a novel to push it forward.
Uh, you can't escape the implied inflation wealth tax by going to a different country, while you could escape a wealth tax by doing so. [Edit: Oops, already said]
Having said that, I agree with you that at .5% it wouldn't make much of a difference, though Graham might be right that even that little is still enough to start people thinking about changing their behavior due to the tax. Also Elizabeth Warran's implied 6% on billionaires accounting for the extra amount charged to cover her healthcare plan would have definitely driven people who were expecting to ever get huge startup wealth away.
Thanks for those examples. I have been looking for cases of movies also. Also it is good that you had here an example of something that a lot of people would view as a negative case (making the invention of the hydrogen bomb faster).
What surprised me and conflicted with my intuitions is the way that works of art pushing already highly familiar ideas that already had lots of artistic works about them are capable of still having a huge effect if they catch the public imagination in either a way previous works hadn't, or that this particular generation of movie goers hadn't been affected.
Obviously The Day After and The Holocaust were not the first movies about those subjects, nor even the first hugely popular and successful movies about those successes (or even in the case of The Day After the first movie that is credited with substantially improving popular awareness on the subject). But despite the fact that it would seem like something which had already been done, there seems to be a clear argument that each had an important effect on the margin.
I'm pretty sure it is actually the same case with the classic slavery example of Uncle Tom's Cabin. I mean, I don't know much of anything about the history, but on reflection it would be very surprising to me if it was the first popular novel focused on the theme of slavery being terrible. And there had at that point been a century of abolitionist activity as a central theme of political life. But it still plausibly had an important marginal influence.
This makes me update away from my view that writing books pushing specifically an AI safety angle wouldn't be useful because it has already been done and people are aware of the ideas. Though I still think that ideas about how to make sure that there is a decent distribution of resources that can make a post human labor society an actually good thing for almost everyone are far more neglected.
Thanks, that's brilliant, and gave me several new ideas on keywords to look for.
The tech founder liquidating his holdings to solve homelessness may or may not be a good idea, but it is not a bad idea because it would suck resources dedicated to tech out of the economy. He'd sell his shares to other people, who now own the tech stream of income, and then take their money and use it to solve homelessness. The only possible economic downside is a bit of inflation as the implied velocity of money goes up slightly since most capital owners spend a very small portion of their wealth each year, while he presumably would be spending it faster. But a 20 billion or even 50 billion one time spend in terms of that is not even a rounding error to the total economy.
Based on my priors about how groups like the FDA and CDC work that seems unlikely to be true. My strong impression is that the system is designed to minimize the odds of things going wrong in a way that will generate headlines based on errors of commission.
Is there any source showing how the decided on this trial duration as something that would balance the risks involved with a deadly epidemic?