ChatGPT
+1

Sep 7, 2026
•
8 min read
"ChatGPT uses Google or Bing" got corrected this week, and then argued with. The Peec AI team documented ChatGPT's own family of indexes: web, PDF, YouTube, news, Wikipedia, local, finance, legal, medical, shopping and more. The evidence runs from sworn court testimony to job postings, from live A/B tests to a billion-page crawl experiment. But the objections underneath the post are serious too, and almost nobody read them. I read both.

ChatGPT
+1

Sep 7, 2026
•
8 min read
"ChatGPT uses Google or Bing" got corrected this week, and then argued with. The Peec AI team documented ChatGPT's own family of indexes: web, PDF, YouTube, news, Wikipedia, local, finance, legal, medical, shopping and more. The evidence runs from sworn court testimony to job postings, from live A/B tests to a billion-page crawl experiment. But the objections underneath the post are serious too, and almost nobody read them. I read both.

ChatGPT
+1

Aug 17, 2026
•
9 min read
In short: three names this week (Suganthan Mohanadasan's research, Rand Fishkin's video, Dixon Jones's commentary) all pointed at the same thing and corrected the most expensive misunderstanding in GEO. Suganthan read the network traffic across 60 conversations: ChatGPT writes brand names into its own first search query before it fetches a single page. Those names come from the model's memory, not the web. A brand that ChatGPT names in its own query gets mentioned in the answer 68.9% of the time; a brand that's only fetched but never named, 2.1%. About 33 times the difference. So the fan-out isn't a search for candidates, it's the model walking down a list it already had. Here's what that means, plus the one test to run this week.

