[charcircuit]

>That seems pretty broken to me for the biggest search engine on the internet.

Not really. There isn't much value to be gained in ranking and sorting a bigger top N of the 8.5 billion results. It is better to encourage people to refine their queries to something less general.

[christoph]

I'm sorry, I don't like to argue on HN, but this is a really silly thing to say.

For instance, search for: "how voltage affects metal in a frozen liquids"

About 15,100,000 results (0.41 seconds), but you can't get past page 10. Page 10 of about 92 results (0.44 seconds)

And on page 10 is a result with this as it's description:

"lyrics to let it go. wolf pup for sale. the taste of chicago 2015 hours. crash landing auto water. prefect des yvelines france. 90s music hits playlist."

Sorry, but it's broken. You can argue until you're blue in the face, but it didn't used to work like this.

[reportgunner]

The amount of results is probably made up at this point.

[eddythompson80]

> It is better to encourage people to refine their queries to something less general

My search experience has been consistently the opposite. Google rejects more defined queries. The more terms you add to your queries, the more likely it’s that it’ll just ignore them.

This is a recent trend btw. It wasn’t that use 4 or 5 years ago, but it’s certainly increasingly random the more terms I add to my search query.

One of my biggest pet peeves with ChatGPT is how conversational I have to make my query to be. I can’t just type terms and let it figure them out. But that’s to be expected I guess. Google search seems to struggle with both atm. I have no doubt they are aware and working on it, but it just sucks atm

[drpixie]

I've observed the same. And the last few results might as well be random - no clear relationship to my search.