You are viewing a single comment's thread from:

RE: Lessons Learned from Curation Rewards Discussion

in #steem9 years ago

With the advent of advanced AI algorithms it is clear that computer models can provide high quality estimates of the value of a post and do so faster than any human could.
Assuming we have the right incentives, the participation of bots is a good thing.

The problem with AI bots is that you don't know what they'll learn. Remember the Microsoft chatbot that turned into a Nazi within a couple of days?
Voting bots would learn to vote for posts with high rewards, which could mean posts for which many bots vote - not necessarily posts with good content.

Sort:  

This is exactly the challenge we face. When we reward voting with a share in the result, voters will start predicting voters in a circular prediction. With enough "predictable agents" in the system the outcome is everyone votes on one thing.

NSFW content for example

Ha, I just Googled
Microsoft Nazi AI; pure class :)