Comment by @punkess • Hey

do you think whatever protocol meta or google or any other „big enough“ (web2) company releases will get the requirered trust to build a global base layer

Stats

Comments

Here's the Ben Thompson article I was referencing: https://stratechery.com/2023/ai-and-the-big-five/
I think they would, yes. "Trust" in an open protocol wouldn't be required, it would be open. But I actually doubt its release. We're about to watch one of the greatest games unfold over the rest of this decade. Every big tech company has incentives to both adopt and reject web3 protocols. Every company's moves affect the rest. And AI disrupts everything. As an interesting example of how many moving parts there are to all of this I'll point out something Ben Thompson talked about today: Facebook ads got killed by Apple's new privacy policy. Facebook ad algorithms were generated using CPU compute. Instagram Reels algorithms used CPU compute but now they need prediction models to grow, which use GPU compute. Meta needs prediction models for content generation in VR, again GPU compute. The company needs massive GPU power which is perfectly aligned with AI prediction which is perfectly aligned with having the best advertising business on Earth. I don't think they'll be able to drop the addiction to advertising dollars, especially if it increases in effectiveness, which means there's little incentive to open up their platform to everyone else. It's going to be incredible to watch how this all shakes out.