>We don't have the same problem of discerning intent like AI would. The same problems? No. But humans pretty notoriously don't always communicate effectively.

>We don't have the same problem of discerning intent like AI would. The same problems? No. But humans pretty notoriously don't always communicate effectively. >It's pretty obvious when someone is communicating nothing but positive things about a product, or nothing but negative things about another competing product. Sometimes it's more subtle... but seriously, if you read enough comments you will locate the ones that are obviously shilling or astroturfing. Yeah, you'll notice the ones that are obvious, but that doesn't really tell you much about your overall accuracy, since you don't know how many non-obvious ones you're missing.