Kagi AI generated summary. (I had to do it)

The web has become an extraordinary public resource, but it is now at risk of being destroyed by the advent of AI. Generative AI models like large language models (LLMs) are disrupting the traditional relationship between writers/creators and their audiences. LLMs can synthesize answers to queries, cutting out the original creators and leading to the rise of “large language model optimization” (LLMO) - manipulating AI outputs to serve special interests. This threatens to degrade the quality of information on the open web, as creators may stop producing content for the public commons. To preserve the web, search engines need to act more like publishers, platforms need to nurture human creative communities, and AI developers must recognize the importance of maintaining a healthy web ecosystem for their own benefit.

Exactly! We need HIPPA style laws covering all personal data.

So it doesn’t actually matter if Kagi can do something SearXNG can’t.
It’s cheaper. No argument with that.

Because it’s not just that.

So I only have to find one thing Kagi can do, that SearXNG can’t?
Then what?

• Link
• Lynk
• Lync
• Linker
• Linkist
• LinkPage
• PageSave
• PageSaver
• PageList
• SavePage
• WebMarks
• Webory (WebMemory)
• Webist

Just some quick ideas. Some may be used already. I didn’t bother checking.

• Netmarks

It always seemed like a massive amount of money and effort, to replace a couple dozen low wage workers. In the end it didn’t even do that.

Notification controls on android are pretty great in my experience.

Most apps (good ones anyway) breakdown different types of notifications, and you can turn off the ones you don’t want. And if they don’t, you can just turn off all notifications for that app entirely.

It all works pretty well.

“It came up with more or less the same recommendations. Though it didn’t fully understand the specific target goals of your project, so our recommendations are more complete and actionable ready.”

What you just described is roughly exactly how the internet actually does work. There are some additional details as you suggested, to make it usable. But that’s the basic foundation.

Until their users found out.

Even an audit is really just someone else saying “trust me bro”. You have some level of trust.

My confidence is inversely proportional to any evidence you have to the contrary.

They don’t keep a search history. Your searches aren’t tied to anything, because they aren’t even saved. They don’t have any reason to save them since they aren’t selling any targeted ads.

That’s why all the top results on every search are actual results, instead of a bunch of ads.
Totally worth it.

Kagi does it already.

!Lemmy [stuff]

Gets you stuff from all over Lemmy

Reasons and justifications are different, and largely unrelated.

Nobody was talking about justifying anything, just the reason.

I never actually asked for a reason. I only pointed out that ability isn’t a reason.

You might find more productive meaningful exchanges, when you engage with actual people rather than your own straw men.

But maybe meaningful dialogue isn’t important to you. In which case I have nothing more to say.

Being able isn’t a reason to actually do something.
A reason would be “Because I feel like it”

That impressive then. I don’t know.

Can you Boost, favorite, mention, or direct message them? No not really. It’s a very limited subset of the Mastodon experience. They have vastly different feature sets, mostly without corresponding analogs. That’s not remotely what I would call seamless.

Or maybe it literally is seamless. There are lots of missing and incomplete pieces that would need to be stitched together, but aren’t.