I just don’t trust anything that an AI Techbro Slopper company says or does. You mentioned their privacy policy? Well anybody who ships AI is sending your information to an AI Datacenter which inevitably ends up in the hands of both advertisers and fascists.
Of course if we’re comparing Kagi to other searches then they’re still better, and that has value, but it’s not for me.
My beef isn’t with and has never been with the concept of a datacenter, lol. LLMs are so far removed from just a regular datacenter. They’re a shit smear across human history which will be reviled to future generations. There are few things more evil than a Ponzi Scheme which boils the earth and sells pure disinformation, even just a regular search engine company and advertisers will never measure up to that.
And people aren’t using them as search engines, either. They’re having entire conversations with them, and giving them extensive amounts of contexts in the hopes of “prompt engineering”.
That completely depends on what model you pick for your queries/chats; they have a little ratings system for the speed quality and privacy. There are also zero data retention policies for a lot of providers; I don’t trust that but obviously it is what it is. It’s also worth noting that those features are entirely opt in and customizable. Don’t want it? Don’t use it. You can turn off the widgets and disallow it entirely. You can also try to filter out a lot of AI slop and they also take community reports for slop sites and display AI and privacy risk badges in results.
Um… Okay then. At least I know you’re not an LLM because an LLM would have written that properly…
And yes, Sam Altman is very much an idiot, along with all the other technofascists? I’m not quite sure how I’m getting lumped in with billionaires, but okay.
I do videcode some personal projects, and I do make use of LLMs to some limited extent professionally, but I’m also a perfectly cromulent developer specializing in resource constrained systems programming and have been since long before the transformers paper dropped… Not sure what that has to do with anything. I write modern C++ a hell of a lot better than frontier models and I explicitly mark my slop projects with a little “hallucinated slop” badge. If that makes me Satan to you, then okay, I guess; still not too sure how exactly my background and personal preferences have anything to do with a paid search engine… But you do you.
My apologies I assumed this response was in a thread on another post about self hosted models. I will not reread your comments and formulate a new response, because you’re still an incompetent slopper and that would waste everyone’s time (not that your time has any value).
I understand. No biggie. I’m used to dealing with LLMs; just like you, they’re often confidently wrong and fundamentally incapable of reason. We can move on.
I just don’t trust anything that an AI Techbro Slopper company says or does. You mentioned their privacy policy? Well anybody who ships AI is sending your information to an AI Datacenter which inevitably ends up in the hands of both advertisers and fascists.
Of course if we’re comparing Kagi to other searches then they’re still better, and that has value, but it’s not for me.
You DO realize that all search queries that aren’t on your own self-hosted searxng or something go to a datacenter, right?
Of course your self-hosted searxng will still route queries to multiple different datacenters, so in reality you’re going to need your own index
My beef isn’t with and has never been with the concept of a datacenter, lol. LLMs are so far removed from just a regular datacenter. They’re a shit smear across human history which will be reviled to future generations. There are few things more evil than a Ponzi Scheme which boils the earth and sells pure disinformation, even just a regular search engine company and advertisers will never measure up to that.
And people aren’t using them as search engines, either. They’re having entire conversations with them, and giving them extensive amounts of contexts in the hopes of “prompt engineering”.
That completely depends on what model you pick for your queries/chats; they have a little ratings system for the speed quality and privacy. There are also zero data retention policies for a lot of providers; I don’t trust that but obviously it is what it is. It’s also worth noting that those features are entirely opt in and customizable. Don’t want it? Don’t use it. You can turn off the widgets and disallow it entirely. You can also try to filter out a lot of AI slop and they also take community reports for slop sites and display AI and privacy risk badges in results.
A statistical model which could theoretically peak at 94% accuracy to a random human with infinite training will always be incompetence.
You are incompetent, slopper. You and Sam Altman are both idiots.
Um… Okay then. At least I know you’re not an LLM because an LLM would have written that properly…
And yes, Sam Altman is very much an idiot, along with all the other technofascists? I’m not quite sure how I’m getting lumped in with billionaires, but okay.
I do videcode some personal projects, and I do make use of LLMs to some limited extent professionally, but I’m also a perfectly cromulent developer specializing in resource constrained systems programming and have been since long before the transformers paper dropped… Not sure what that has to do with anything. I write modern C++ a hell of a lot better than frontier models and I explicitly mark my slop projects with a little “hallucinated slop” badge. If that makes me Satan to you, then okay, I guess; still not too sure how exactly my background and personal preferences have anything to do with a paid search engine… But you do you.
My apologies I assumed this response was in a thread on another post about self hosted models. I will not reread your comments and formulate a new response, because you’re still an incompetent slopper and that would waste everyone’s time (not that your time has any value).
I understand. No biggie. I’m used to dealing with LLMs; just like you, they’re often confidently wrong and fundamentally incapable of reason. We can move on.