# Precommercial scitech funding is taste based By definition precommercial scitech can’t be judged on the ür-metric: money (or more nebulously, predicted value). If you could, it wouldn’t be precommercial, it would just be a high-risk investment. Without money as a metric, each project needs to have a custom definition of success. Is it pulling all the carbon out of the atmosphere? Enabling people to travel from one side of the world to the other in an hour? Discovering a fundamental truth about the universe that doesn’t seem to affect anybody’s life? Creating a new field? A utilitarian calculus might claim to be able to objectively compare projects based on something like “minimizing suffering” or “maximizing flourishing” or “expected GDP in the year 2200” but the choice of that metric is ultimately based on taste. Not to mention that any algorithm that converts between actually measurable things and those metrics has a lot of taste built into it. Why do we think LIGO is valuable? Because a large chunk of physicists say so, and their opinions seep into other people, and we take either the opinions of some subset of physicists or people they’ve influenced opinions seriously. Previously a large chunk of physicists thought differently. Did anything change about the value of discovering gravitational waves or the chance of it working? No. Taste changed thanks to a combination of persistence of a core group + changes of who was in power + luck -> results that started to change perceptions. Taste dominates when there’s no objective measure or agreed upon definition of what ‘works’ means. On the extremes, there are projects that have no clear definition of what works means or will only work on large timescales. How do you know when you’ve founded a new field? What if the work you do now is ultimately incredibly useful for a project that we can’t even imagine yet? [[It is impossible to derisk an idea when there is no definition of what works means]]. Taste still dominates even in situations when you can ‘run the numbers’ ([[Startup by spreadsheet]]) to make an argument for a certain narrative. Taste sneaks in both because the numbers have assumptions built into them and because without a narrative that ‘this will make us a bunch of money’, taste determines what narrative is compelling. This all isn’t to say that we should just give up and fund things randomly ([though some studies suggest it might be more effective than other systems)](https://nintil.com/funding-lotteries/). My argument is simply that we should look the taste-based nature of this world straight in the eye and think seriously about it instead of pretending otherwise. ### Related * [[Value capture is a utilitarian fantasy]] [Web URL for this note](http://notes.benjaminreinhardt.com/Precommercial+scitech+funding+is+taste+based) [Comment on this note](http://via.hypothes.is/http://notes.benjaminreinhardt.com/Precommercial+scitech+funding+is+taste+based)