Will John's air conditioner test find at least 50% better temperature delta with two hoses compared to one hose?
54%
chance
Details on LessWrong: https://www.lesswrong.com/posts/AMmqk74zWmvP8tXEJ/preregistration-air-conditioner-test
Update: took a first pass at this today, experiment ran into some issues (including one issue which was pre-registered as an invalidate-and-retry condition). Details and current planned changes for the next attempt are here: https://www.lesswrong.com/posts/puv8fRDCH9jx5yhbX/johnswentworth-s-shortform?commentId=AeKtgW5JWWaK8qfdJ
KD89042 is betting YES at 50% a month ago
isn't john the one running the experiment? it seems like he'll bestrongly motivated to avoid issues with the cardboard hose, and in any case will be pricing that into his 90%
Stephen Malina bought M$10 of NOa month ago
I feel like people are really neglecting the cardboard hose component here. Even if John's arguments seem much more trustworthy, I put decent probability on a janky cardboard hose setup causing efficiency loss in itself.
KD89042 bought M$23 of YESa month ago
idk anything about ac but it seems like swentworth has put a lot more thought into this compared to christiano; if he thinks it's 90% then its probably a buy at 55
I just added M$ 50 of liquidity. Can't wait to see the results
Undox bought M$20 of NOa month ago
I am not an engineer or physicist. I reckon the single host has a compensating advantage. The temp difference between the hot part of the heat exchange and the passing air will be higher for a single hose. Lets say outside temp is 40c, inside is 25c, maybe heat coil is 60c? then there is a 35c difference instead of 20c. So here it will be more effective.