I haven't asked Google a question it has failed to provide a more than adequate answer to in ... months? years?
And on all my devices, I run google search with &udm=14, so I am not talking about AI summaries. I also have search personalization disabled.
I see a lot of people complaining about this on HN. It simply doesn't match my experience at all, in any way.
It's easier to add Reddit at the end to get a more accurate question repeated, and skip the sponsored SEO crap.
And a list of links to original sources or close to it is precisely what I do want.
If you want an LLM to generate an answer from its training data, that's fine, but go use a different search engine instead of demanding that the one many of us have relied on for decades has to do that.
And I actually agree with the last point. While there are entire categories of questions that I now prefer an LLM to to any search engine, when I want a search engine, I specifically do not want LLM summaries, which is another thing I like about Kagi: they allow me to choose when I want to see an LLM summary and to turn off summaries altogether.
(this is really not meant to be an ad for Kagi, I presume that most HN users are familiar with it already and don't need yet another random endorsement, but I honestly don't know how to talk about my experiences with search over the past several years and my dissatisfaction with google without talking about it)
firefox: keyword "g" configured as google search plus &udm=14
firefox: keyword "a" configured as google search without &udm=14
works for me!
So, if not that, then what?
The AI summary is not the problem; you could take it away and the experience would be just as poor.
In fact, the AI summary slightly improves the experience for faster readers.
Google searches don't produce good results these days. The enshittification has become too extreme. Google openly admits as much (and further intensifies the enshittification) by placing a huge AI summary above those results.
The answer is self evident. If, before, you were relying on clicks resulting from google searches, today you need to be what an AI recommends when somebody uses an AI like they used to use google. (Users will eventually become more sophisticated though!) Lots of people are using AI like a search engine and getting better results than google gives simply because massive resources are currently being put into training AI, while mere neglect is insufficient to explain how fast Google search results are getting worse.
Is this how AI companies plan to cash in? Accept money from advertisers to promote their products in interactions with their LLM's? Were I an advertiser, I'd be trying to get Anthropic to take my money instead of giving it to Google. AI might be what finally makes it impossible to tell content and ads apart. That's great for advertisers... I guess. Not so great for the rest of us.