wujerry2000 a day ago

Running a test with browser agents and asked it to leave a comment on a Medium article.

Came back 1 month later and found it had risen to be the most liked comment on the thread.

Is there research around large scale Turing Tests like this? Using human environment signals as research indicators? (safe/ethically of course)

I wonder if its possible to train an agent to become an expert at going viral on HackerNews using upvotes as a reward signal.

Maybe this experiment is already running...

  • dtagames a day ago

    LinkedIn has been taken over by this. Many (most?) viral posts and comments are from bots now. Not great.