We spent ~20 person-months or something on Foretold.io, and have stopped working on it.
The main reasons were:
Metaculus seemed to be growing & hiring a lot. (Originally this seemed less likely)
It was more work than expected. I originally was hoping we'd find some clever ways of making it much simpler, we didn't really.
In order to keep it active, it would have taken ~2 full-time people, at least (one programmer, one community organizer/question writer). This was a lot for the time.
We weren't really able to solve the problem of what to do about data management. (Large tables of semantic predictions).
We moved Squiggle into a separate project, which ideally could be used with Metaculus and other platforms.
I could see writing this as a ~2k word blog post, but I'd expect this to take a week or so.
We spent ~20 person-months or something on Foretold.io, and have stopped working on it.
The main reasons were:
I could see writing this as a ~2k word blog post, but I'd expect this to take a week or so.