I propose the following idea on how to increase probability that artificial intelligence will play out well. The idea is to start an economic movement like the one I have described in the following post. The problem that it's supposed to solve is that there are certain things that needs to be done with regards to artificial intelligence risks that are poorly incentivized. Links (both posts describe the same idea, I recommend to read the first post): 1. [[Effective altruists would be more effective if they did this]] 2. [[Alignment between humans/Cooperative equilibrium/Cooperative equilibrium|Cooperative equilibrium]] Here's what we would have to do start it: 1. Ideally, find some investor/funder that will provide funding for this project. I have a list of potential funding opportunities that we could try. 2. Record a YouTube video that: 1. Explains the idea of the economic movement. 2. Announces that we will retroactively reward people (by giving them money) who did something good and were not sufficiently rewarded through traditional market incentives (including the people who rewarded those people), or that we will donate money to projects that need money to do something good (most likely non-profits because in case for-profits the traditional market offers incentive to invest in them), specifically something that counters AI risks like AI-enabled biological weapons. Who gets rewarded can be decided by me, you, investor/funder and/or someone else... 3. Do what has been announced. 4. Create a decentralized prediction market to increase observability - to make it more known how effective a given project is, how much reward it deserves. In this decentralized prediction market, people can bet on certain questions, including those questions that are related to how effective a given project is. I have some other ideas as well. %% // question: will we be able to do point 3, given that we probably don't have info about what people did %% %% Two solutions: 1. talking to specific non-profits to make it possible to prove that someone donated money 2. decentralized prediction market 3. off-loading enforcement to the future %% %% # Second idea I'm interested in game theory a little bit. I'm not an expert, but I watched a course on Coursera and explored game theory a little bit, so I'm familiar with some concepts. I was thinking a lot about how to apply game theory to predict what will happen in the future regarding artificial intelligence, and then figure out what to do about it. I discovered that the textbook game theory is a little limited, and relies on unrealistic assumptions. %% %% Convince people, especially influential people like politicians, executives and employees of AI companies to act ethically. I believe that there is a non-obvious reasoning that leads to a conclusion that they have a higher interest in acting ethically and cooperatively than they believe. The idea is to popularize that reasoning. I would tell you what the reasoning is later. // read that again, and think if it makes sense %%