I would like for Kagi to collect data on my interactions with the platform, and leverage this data to deliver user-behavior driven features. User workflow may be impacted by additional privacy related preferences.

As a user:
I want Kagi to deliver an optimized experience.

I love and appreciate the wealth of innovative tools that Kagi makes available to subscribers. I have limited attention and patience budgets when it comes to tweaking and tuning.

Features like ranking adjustment are high-five worthy, but it's a tool to control page rank, not a solution to the core complaint: "Why aren't these results presented in the order I expect/desire?".

As a subscriber:
If I am paying for a service type that is generally regarded as free, it needs to exceed the value delivered by the free service. Kagi is crushing this delivery in several under-served dimensions of it's product class, but it's lagging in several that are driven by data.

As an engineer:
I see this capability as a blocker to delivery of a critical "class" of features that I expect from a modern search provider. I'd expect implementation to be arduous but not impossible. Modern responsible data handling is a robust practice.

As an internet user:
Data is a tool that has been wielded against me by the pointy end of the Fortune list. The value of user data is unquestionable. Why does it always have to be evil? Please provide this type of value in a way that doesn't make me feel ick.

  • carl replied to this.

    maurs but it's lagging in several that are driven by data.

    Can you give examples?

      carl not that I endorse this as a good direction, but just as a strict high level example: the Kagi assistant is not Siri in a number of ways. It not in the same league as other “assistants” from an accessibility perspective, and it has no data to inform personalized responses.

        Here is a real use case: I want Kagi’s help making the page rank feature “just work”.

        What’s the best way to rank these? Man I don’t know. Why don’t you tell me Kagi? Haven’t you been paying attention? You’re the one on the clock here. Do I scroll really far to find stack exchange consistently? Maybe put stack exchange higher up? Quit asking me, I’ll peek under the hood if I feel compelled. Just do better.

        (I apologize for the snark. I’m trying very hard to stay away from technicals and solution-ing. This attitude was a story telling device not an opinion)

        I mean, the reason I pay for Kagi is exactly because the search results by default are good and ranked better than other search engines., without me having to manually rank. If you think Kagi has failed in that function, then there's no reason for you to pay for and use the service.

          carl I have no intention of implying the service Kagi provides is poor - full stop.

          The existence of a page rank adjustment tool suggests that some adjustment is desired. I would like that adjustment to be automatic and based on my previous interaction with the feature, as well as any other pertinent interactions i make with the platform.

            I want Kagi to collect and use less of my data not more.

              No one is typing