But increasingly, discovery seems to be happening somewhere else: ChatGPT Claude Perplexity AI-powered search and assistants

These systems don’t “rank pages” the same way search engines do. They select sources, summarize them, and recommend them directly.

What surprised me while digging into this: - AI models actively fetch pages from sites (sometimes user-triggered, sometimes system-driven) - Certain pages get repeatedly accessed by AI while others never do - Mentions and recommendations seem to correlate more with contextual coverage and source authority than traditional keyword targeting

The problem is that this entire layer is invisible to most builders.

Analytics tools show humans. SEO tools show Google. But AI traffic, fetches, and mentions are basically a black box.

I started thinking about this shift as: GEO (Generative Engine Optimization) or AEO (Answer Engine Optimization)

Not as buzzwords, but as a real change in who we’re optimizing for.

To understand it better, I ended up building a small internal tool (LLMSignal) just to observe: - when AI systems touch a site - which pages they read - when a brand shows up in AI responses

The biggest takeaway so far: If AI is becoming a front door to the internet, most sites have no idea whether that door even opens for them.

Curious how others here are thinking about: - optimizing for AI vs search - whether SEO will adapt or be replaced - how much visibility builders should even want into AI systems

Not trying to sell anything — genuinely interested in how people here see this evolving.