Results in a search engine always require scrutiny (whether provided by an AI or not), so I was pondering the difference in how search engine results fail us when they are, or aren't, provided by an AI. After all, somebody might (and probably does) argue that AI search results are no more untrustworthy than their predecessor! But, I think the stark difference is that there is actually something very predictable and comprehensible in the way that search results have historically been wrong: They are completely submerged in their source context (and as far as mainstream AI goes, erasing this context in the name of convenience is the very nature of what's going on). Each site that offers me results in a "traditional" search as we know it conveys intent: A technical article from a startup is going to serve recruitment ads, but they have an incentive to serve me quality. A wikipedia article might be target of history revisionists, so this is where my attention could be. A fresh stack overflow post will be swarming with gamification-reward-addicted posters that will serve low quality advice, I will be wary. There is a skill in navigating this, it can be learned!
AI results however? It's me against a black box - likely operated by profit motives - now. How am I supposed to reason, to retain my grip on reality in this?
