Comment by kevin_thibedeau

That matches my experience. Kagi doesn't surface much content beyond what Google/Bing do. What it does better out of the box is guessing which content is low-quality and displaying so that it takes up less space, allowing you to see a few more pages worth of search results on the first page. And then it lets you permanently filter out sites you consider to be low quality so you don't see them at all. That would have been awesome 10 years ago when search spam was dominated by a few dozen sites per subject that mastered SEO (say expertsexchange), but it is less useful now that there are millions of AI content mills drowning out the real content.

For content that isn't time sensitive the best trick that I have found is to exclude the last 10-15 years from search results. I've setup a Firefox keyword searches[1] for this, and find myself using them for the majority of my searches, and only use normal search for subjects where the information must be from the last few years. It does penalize "evergreen" pages where sites continuously make minor changes to pages to bump their SEO, which sucks for some old articles at contemporary sites, but for the most part gives much better results.

[1] For example: https://www.google.com/search?q=%s&source=lnt&tbs=cdr%3A1%2C...

I use Kagi personally every day and my results are definitely not full of AI garbage so would like to better understand your context.

Have you reported any of those issues to Kagi (support/discord/user forum)? We are pretty good at dealing with search quality issues.

The UK NHS website is usually pretty good for this so sticking "NHS" in the search terms might help, although I imagine they may not cover non-UK brand names.

> I ended up ditching search alltogether and ended up finding a semi-relevant paper on the nih.gov and going through the citations manually to trying and find information.

I've been doing this for years now. The normienet as I call it is nigh worthless, and I don't even bother trying to find information on it.

I also use it daily. One of my favorite functions is being able to boost certain domains and block or downgrade results from other domains. So I boost results from domains I trust which significantly improves my results. They have a page with commonly boosted/blocked/downgraded sites which serves as a good starting point.

I'm a little sad for anyone who didn't get to experience the Internet of the twentieth century. It was a unique point in time.

I'm ready to pay for a walled garden where the incentives are aligned towards me, instead of against me. I know that puts me in a minority, but I'm tired of the advertising 'net.

I only just put it together but Peter Watt's Rifters series is some epic earth grimdark hard-sci-fi, the first taking place as practically horror, confined deep under water.

But my point is, the latter books have this has amazing post-internet, just a ravaged chaotic Wildlands filled with rabid programs & wild viruses. Packets staggering half intact across the virtualscape, hit by digital storms. Our internet isn't quite so amazing, but I see the relationship more subtly with where we have gone, with so so so many generated sites happy to regurgitate information poorly at you or to sell you a slant quietly. Bereft of real sites, real traffic. Watts is a master writer. Maelstrom.

First book Starfish is free. https://www.rifters.com/real/STARFISH.htm

> It really is a werid feeling remembering the internet of my youth and even my 20s and knowing that it will never exist again.

User facing ability to whitelist and blacklist websites in search results, ability to set weights for websites you want to see higher in search results.

Spamlists for search results, so even if you don't have knowledge/experience to do it yourself, you can still protect them from spam.

It's recreation of e-mail situation, not because it's good, but because www is getting even worse than e-mail.

A mesh network on top of IP with an enforcable license agreement that prohibits all commercial use would suffice to get the old net back. Bonus points if no html/css/js is involved but some sane display technology instead.

The old internet seems to be doing much better. But it lost most of its users in the last 15 years..

What do you mean by this? How do you find the old internet?

Reddit is already lost. I was talking to the mods in a large political subreddit and they said after Reddit started charging for API access, all the tools they used to keep on top of the trolls and bots stopped working, and the quality of the whole subreddit declined visibly and dramatically.

If you know anyone who works in marketing/PR, ask them how they use Reddit. That has been gamified as much as SEO since about 2020. I’m assuming, anything except “why is there a fire in this street?” kind of posts are just ads at this point.

It's also not much use to anyone who doesn't use Google ever since Reddit started blocking all crawlers besides Googlebot. Old cached results might still show up in Bing/DDG/Kagi but they can't index any of the newer stuff.

Most of the Reddit content is now[0] fake.

[0] Gradually for several years already.

It's surprising how many times you see this pattern on HN

"Google sucks!"(50 upvotes)

"That's why I use Kagi!"(45 upvotes)

"Actually Kagi has the exact same problem and you have to pay for it."(2 upvotes)

Search “peachick”, it works fine. I assume Google would be the same.

I guess using the correct terminology matters.

Kagi's images are an entirely different set for me.

Your unauthorized access has been reported to the Fair Use Bureau.

There's ways to do this without destroying anonymity. Ideally, you verify you're human by signing up for some centralized service in real-life, maybe at the post office or something. And then people can ask this service if you're real by providing your super-long rotating token. So, just like an existing IDP but big.

That's how the Internet works in China.

Couldn't reproduce - in fact, the second hit is a threads version of the same post - but I get no AI suggestions for this query. Humorous Google queries (or AI queries more generally) are definitely a trope, so I can never really tell if they actually happened or if it's all for karma.

This peacock thing was the last straw for me. I installed Kagi just moments ago.

And of course the first image for "baby peacock" is the same white chick thing… obviously because this story is making the rounds —_—

AI tools on the search page: sure, cool. I use perplexity a lot, actually. I'm in favor of this.

Search results that are full of content mills serving pre-genned content: no thanks. It's in the same category as those fake stackoverflow scrape sites.

Not sure if you’re being sarcastic, but they’re not talking about AI features of the search engine itself (Kagi has those too), but about nonsensical AI generated content on the web that exist solely for the purpose of getting you clicking on some ads. Kagi tries to make those sites stand out less on the search results.

Perhaps you're thinking of the Wikimedia Foundation.

There is plenty of space there for more volunteer editors to verify content, and likewise, WMF operates its own cloud platform where developers are automating tools that do maintenance and transformation on the human-contributed content.

Then, there is Wikidata, a machine-readable Wiki. Many other projects draw data from here, so that it can be localized and presented appropriately. Yet, its UI and SPARQL language are accessible to ordinary users, so have fun verifying the content there, too!

This issue in terms of cost is that if you want this to be truly human-verified for real, you're gonna have to dip into the real world.

Revival of the curators.

How can that ever work in a world filled with people that are eager to lie to you?

In the world of content moderation, we refer to this as constructive friction. if you make it too easy to do a thing, the quality of that thing goes down. Difficulty forces people to actually think about what they are writing, whether it is germaine and accurate. So generative AI, as you point out, removes all the friction, and you end up with bland soup.

You would have to be soft in the head to rely on any LLM for researching information on a medication you're actively taking.