In the canonical story of the playpump, we had an inventor who imagined the future benefit of his dual use water-well playground. He failed. We learned that we waste money when we rely on how well we imagine our ideas will work. We learned that we need to build and measure the results.
Those of us who speculate on their efficacy of curbing extinction risk and those of us who speculate on their improvement of future well being are all in the same boat as the inventor of the playpump.
I don’t think this is the greatest analogy, but I do agree that there are huge risks in imagination. I would say EAs failure (in my mind) through funding Open AI and betting the house on alignment research could almost be considered kind of “play pump-esque” examples. Good ideas coming from the best intentions, but without past evidence as a rock to stand on.
The debate sets aside a founding principle of EA.
In the canonical story of the playpump, we had an inventor who imagined the future benefit of his dual use water-well playground. He failed. We learned that we waste money when we rely on how well we imagine our ideas will work. We learned that we need to build and measure the results.
Those of us who speculate on their efficacy of curbing extinction risk and those of us who speculate on their improvement of future well being are all in the same boat as the inventor of the playpump.
I don’t think this is the greatest analogy, but I do agree that there are huge risks in imagination. I would say EAs failure (in my mind) through funding Open AI and betting the house on alignment research could almost be considered kind of “play pump-esque” examples. Good ideas coming from the best intentions, but without past evidence as a rock to stand on.