Imagine trying to throw a party but half your guests are suddenly like, “Nah, I’m good; I’ll just binge-watch old sitcoms instead.” That’s basically what’s happening with Google’s DeepMind.
During a riveting DOJ antitrust trial (because who doesn’t enjoy a little courtroom drama?), it was revealed that a staggering *80 billion out of 160 billion* training tokens—basically, snippets of content—were tossed aside because publishers said, “Thanks, but no thanks!”.
This isn’t a one-size-fits-all situation. DeepMind’s models have to play nice with the rules set by publishers. If they opt-out, their content is off-limits. But if you think that means the search team is sitting idly by, think again! DeepMind VP Eli Collins confirmed that the search org can still do its thing with whatever’s left, even if some publishers are sitting out.
So, what does this mean moving forward? Is Google’s AI still going to shine, or are publishers effectively dimming the lights? Only time will tell, but it’s clear that in the wild world of AI, every opt-out feels like a little bit of a party foul.
What do you think? Should publishers have more say in how their content is used? Or is it all just a digital free-for-all? 🤔
Leave a Reply