There’s a rush now at Google and Microsoft to incorporate chatbots into their search engines, which seems apt to undercut the usefulness of search engines.
Suppose you want a list of Ruben Bolling’s God-Man comics. The words that a large-language model will string together about God-Man, based on having taking the whole internet as its corpus, will be nonspecific, incomplete, and probably error-ridden. The best source on this topic would be an official webpage or a fan page. There is no official God-Man episode guide, but I happen to maintain what is probably the only God-Man fan page. The best answer to the query is not to digest my page along with ten thousand others, but just to link to my page.
For lots of perfectly ordinary queries, one wants a dedicated webpage or site rather than the word-salad gestalt of a general purpose tool. So a traditional search engine can provide answers that a chatbot will only obscure.
This post was prompted by OmnipotentFriends, the latest adventure of God-Man and the cause of an update to the fan page.