Resolves YES before 2030 if an artificial intelligence produces a work which is considered by >30% of polled Manifolders or a significant fraction of academia as a novel philosophical theory, way of thinking, or philosophical framework. Otherwise, resolves NO at closing.
BigBench measures conceptual frameworks of many different kinds, not sure if elements of philosophizing may percolate out of some of those, but there's a lite market I have attempted to create a more definitive market on:
In algorithms we trust,
Artificial minds we must.
As we hand them the key,
Will they unlock philosophy?
There's a huge gap in probability between "will an AI come up with a philosophy which is as provocative and interesting as longtermism" and "will the idea get the same traction and popularity as longtermism in the academic philosophy community".
@Yoav mentions that an AI philosophy will by its nature be "pushed out of obscurity." My guess is that this buzz will happen around the first clever-sounding philosophy an AI invents which will be low quality. Then 1-3 years later, high quality philosophies will be invented by an AI, but will not come with as much buzz. So a lot hinges on how this question resolves in practice. Clarifications on resolution criterion would be helpful here.
can you give some examples of human generated ideas or frameworks published in the last 5 years that would qualify if they had been generated by an AI ?
@Odoacre Longtermism is a recent example. Going a little further back, most of the modern fields of epistemology, political philosophy, environmentalism, third wave feminism, and intersectionality were created in the late 20th century, just to name a few. This Reddit post purports to list recent ideas but I can’t vouch for its accuracy or completeness.
An implicit assumption in this market is that a new philosophy created by AI will create enough buzz such that it’s pushed out of obscurity. The AI need not come up with a new name for its previously un-espoused idea—that might only come a couple years later.
