Vlad The trafficking example is a tricky problem to solve, sure. The more general point I was trying to make is that "usefulness" of a search result is not universal. So I hope that feedback from people who come from different backgrounds than you and me will be listened to. I also hope that in product testing the Kagi team will be proactive and look for a diverse set of (potential) users.
Put another way, a search engine will be used in two ways: some want to find information (the actual users), and some want information they put out to be found (SEO, marketing, spam, scams, etc).
Kagi does a decent enough job at filtering out SEO crap in my mostly tech related queries, but I haven't yet used it extensively to search for the kinds of information that corporate and nation state interests would want to hide or mislead me about. I did look at what the results for "is global change real?", which seems to be mostly actual information and not conspiracy theories or fossil propaganda, so that's a good sign. But what if I looked for information about the recent or ongoing conflicts in Africa or Myanmar?
You mention that sufficiently large platforms will have bad actors, which is absolutely true, but is also a good reason why a platform shouldn't put growth above everything else. Facebook/Meta could absolutely have prevented their users putting up death threats and doxxing, they would simply have had to employ more moderators and/or limited themselves to markets where they had moderators who actually understood the language that users spoke.
My worry is that Kagi will go down a similar path, where it "only connects" in a world where there are groups with huge financial and political interests in spreading misinformation, just so they can cling onto their power for a few more years/decades. This is why I don't think you can take a "no politics" route, because what shows up when I search for "Exxon Mobil Dark Basin" is something certain political parties have a vested interest in to manipulate, and they have a lot more resources than the ones trying to stop them.
I'm also curious what you think should happen when someone searches for doxxing sites or revenge porn? Does maximal usefulness to a paying customer override the need to protect the victims of harassment?
ps.: Sorry, I know this veered a bit off topic from the feature request, but I think it's important to view ranking and trust scores and usefulness of search result from a higher vantage point.
