Right. The myth that keeps getting confidently repeated in HN comment sections is that Mozilla supposedly lost market share due to a series of strategic missteps. But it basically was about the pivot to mobile, and the monopoly lock-in of Google. Actually think one fantastic remedy for Google's search monopoly might be allowing the use of alternative browsers on Android via a pop-up rather than preloading and privileging Chrome. Because browsers and mobile are part of the strategy of creating a path dependency tied to Google search.

But to your point, I think the simple reality is that LLMs are increasingly taking the place of search and so having all your funding based on search licensing might be risky when it's at least possible that we're going to be in a new paradigm sooner than later.

I honestly think AI in the browser right now is generally very half-baked and doesn't have any well thought out applications, and raises all kinds of trust issues. I can think of good applications (eg browse the Kindle unlimited store for critically acclaimed hard sci-fi books), but there might be better ones that I'm not thinking of. It just might make sense to be involved so you went caught flat-footed by some new application that quickly progresses into something people expect. And of course because HN commenters are famously self-contradictory in response to literally everything Mozilla does, it's a damned if they do damned if they don't situation: if they load AI into the browser it's pointless feature bloat. If they don't then they were sitting on their thumbs while the world moved on when they should have been reinventing themselves and finding new paths to revenue.