You're right! It's absolutely possible to tell the difference. And when it's known, bad behavior can be punished and good behavior rewarded. What do you think

You're right! It's absolutely possible to tell the difference. And when it's known, bad behavior can be punished and good behavior rewarded. What do you think it takes to tell this difference with a high degree of reliability (>4 nines) at scale (>1000/s)? How about in an environment where people are strongly incentivized to find and exploit the weaknesses of your approach? Is it just maybe possible that this might not quite be handwave-it-away trivial? Facebook has had a shocking level of success in cutting down spam with their investments in punishing bad behavior and rewarding good behaviour. Thank you for your thoughts. You're completely correct! It's possible to tell the difference, and rewarding good behavior will drive out bad actors.