I wonder if we'll reach a point where YT asks viewers to pass a captcha at the end of an ad to prove that they watched it before getting back to their video.
I wonder if we'll reach a point where YT asks viewers to pass a captcha at the end of an ad to prove that they watched it before getting back to their video.
A: Louis Vuitton
Eventually each ad has an embedded AI that must be run to see the content after the ad. That AI evaluates whether there's a human who looks and behaves like you watching the ad. The endgame requires your AI to do a convincing impression of you to the ad AI, and to hide other signals that would reveal you doing something else, like the sound of a flushing toilet.
When your AI is pushed by more advanced ad AI to be very convincing, it has to start purchasing things occasionally, just like a typical human. It occasionally buys the things in the ads, the way humans are expected to, to avoid revealing it's not you.
This spirals out of control when your AI has to buy more things than a human would, because the ad AIs co-evolve with your AI to expect that. It won't be possible for an ordinary human to watch content unless they run a highly-evolved adblocker AI that buys enough of the things shown by the ad AIs to satisfy the ad company.
Unless WEI becomes a thing.
GPT-4 has entered the chat
Fun way to screw with Google is to pick the worst answer (haven't seen any of the products, worse impression of the brand, etc).
Advertisers are starting to try to measure advertising effectiveness (did the user actually see our ad and like our product) instead of easily game-able metrics (impressions, time on screen, click through).
However, we found that poor ad experiences would result in poor metrics. Advertisers really don't like it when they spend millions of dollars in advertising to get a report that says "your target demographic is less likely to consider your product now after seeing your ads".