Search engines may be on their way out thanks to artificial intelligence, but I found myself wondering: How will AI crawl websites to generate its answers?

I believe that whatever systems companies like Google developed to search and index webpages will simply be repurposed to “train” AI models to judge whether a piece of content is high quality or not. In the end, what’s marketed as AI is often just another algorithm doing what Google’s crawlers have been doing for decades.

Because of this, AI optimization will not be very different from traditional search engine optimization.

Yesterday, for example, I used ChatGPT to look for a good movie. I wanted to start with the earliest classics and work my way through the history of cinema year by year. ChatGPT recommended silent films by Georges Méliès—among the first true movies, dating back to around 1898 (some say 1902). It gave me a link to a free YouTube upload and shared some background information with another source.

But how did ChatGPT know this was a good answer? And how did it know the link would actually lead me to the right place?

The answer is SEO. The information about the film, the link to the video, and the related resources were well optimized and ranked high in Google’s search results. ChatGPT simply relied on that structure to determine that these sources were likely to be high quality.

Of course, you usually have a brief conversation with the AI to fine-tune its responses—this is a form of answer optimization, that is unique to AI because it happens directly between the model and the user. Still, whenever an AI gives me a link, it’s doing so for essentially the same reason Google ranks a link highly.

Therefore, even if search engines fade because of AI, SEO will not disappear. Most of its principles will continue to influence how AI models refer users to your content, just as Google does today. Understanding the basics of SEO is still essential—and worth our time.