Just listened to a really good conversation on @thinkfuture , Chris kalaboukis talking with Andreas, who runs RTOS, about where search is actually headed.
His take on SEO is blunt: it's basically on life support. Search engines are moving toward AI-generated answers, and most content out there isn't adding anything new it's just noise competing for the same generic answer.
What stuck with me is how he thinks about fixing that. Instead of pumping out more AI content (which just adds to the noise), his team actually digs into real conversations on Reddit and Quora to find what people are genuinely saying, then builds content around that. The idea is "information gain" does this actually tell the AI something it didn't already know? If not, why does it exist?
He also made a point I hadn't thought about before: AI tools like ChatGPT hallucinate because that's literally what they're built to do — the "G" in GPT is generative. It's designed to produce an answer, even a made-up one, because that's more "helpful" than saying "I don't know." He mentioned dropping the temperature to zero and getting AI to fact-check itself as ways around it.
The part that really got me was his 10-year prediction. He thinks we're going to move past building apps entirely and start programming hardware instead like being able to tell your home robot to do random specific tasks for you.
Kind of a wild thought, but the SEO point alone is worth sitting with. Feels like a lot of us are optimizing for a game that's already changing.
Anyone else feel like the old SEO playbook is running out of road?
Comments
No comments yet.