Hacker News Reader

Comment by hug | Hacker News Reader

The author would do well to check out Kagi -- the search results for all of the suggested queries are, to my eye, better than the results they found.

Given that Kagi's higher tier plans come with search-enabled LLM chat interfaces, and those searches use Kagi's results (which, again, appear to be superior) it seems to me that you get the best of both worlds: Better search, and better search results to feed into your search-enabled LLM queries.

I am not affiliated with Kagi or anything, it's just honestly that good a product.

View on HN

I second the recommendation for Kagi. It just blows all competitors out of the water in my opinion, and I'm happy to pay for it as long as they keep that quality up.

Third. The ability to uprank sites makes the “finding known content” use case absolutely amazing. The specific postgres docs case in the blog post is one I am constantly using kagi as my external brain for.

Happy customer of Kagi here too. However if I’m being honest, I’m starting to get LLM generated content in Kagi results.

Don’t think it’s their fault or that it happens more than everywhere else neither what they could do about it but it happens.

I am pretty sure 50% of my search results on Kagi (or Google or Bing) are LLM-generated. You can search anything and find a dozen websites that, surprise, have a page dedicated exactly to that topic, organized in three or four neat sections.

The internet is dead and is starting to smell.

Author here. I will be trying out Kagi for some time. Search is worth paying money for. Thanks for the heads-up!

I've been using Preplexity lately while learning to use Yamaha SeqTrack synthesizer. I have the UserGuide, but I find it easier and faster to ask Preplexity for things like "How do I mute a track?".

It never occurred to me I could use Google this way. And it is a novel idea to me, that it seems to be better to use AI than read a manual.

If Google's LLM is trained on the SeqTrack manual, then yes.

If it's not then it will just invent some plausible-souding bullshit that doesn't actually work.

After the fifth time you get burnt by this the whole LLM experience starts to sour.

Yes. I wonder what's LLM providers' approach to ensuring high-quality of training materials?

Lately, I've been more and more disappointed with Kagi. It seems to be falling victim to the same SEO spam. For example, today I was wondering what Scala devs use as alternative to Akka. I searched for "scala akka alternative". Try the same (using Kagi). I get mostly listicles and SEO spam. I have to scroll halfway down the page to see links to "Pekko" and "Zio"... and they are hard to separate from the noise. This is just my most recent example... it just "feels" like this happens a lot more, even with Kagi now.

However, the reason I started paying for Kagi was because they let me completely block websites from search results -- and they still let me do that. That feature alone will keep me as a paying customer for the near term.

To be fair to Kagi:

This was in the top three results, and I can't tell if it's "real", or just a page created to capture those exact search terms!

https://akkaalternatives.com/

If the goal of that site really is just to capture clicks via very specific web searches... and people are creating sites like that at scale... what hope do we have of saving the web? :-(

i would probably pay for kagi if there was a premium tier that excluded LLM features.

Reply on news.ycombinator.com