6

I've tried using the automatic summarization functionality of Kagi a few times now, and each time it's been entirely wrong as well as being slower than reading the source pages.
Whilst I'm sure there are many people who find these sort of summaries useful, they are not a priority for me in a search engine.
If possible, it would be great to have an AI-less plan that was closer in price to the old unlimited plan. If this is not practical, it would be good to have the ability to disable the AI functionality in my account - partly to avoid wasting search credits and partly to allow the Kagi team to gain an understanding of how many users value this functionality.

Example query: "ansible integer value in extra-args"

  • Vlad replied to this.

    theothertom The AI features do not impose any additioanl cost to us (or you) unless you use them. The only thing that is exposed in UI is really Summarize Results which you could easilly hide with Custom CSS.

    In other words there would be no benefit in creating an AI-less plan, just a lot of extra work around billing.

        On the extra cost thing, it may be coincidental timing, but my perception is that the pricing went up as a result of the AI features since they launched together.
        The other aspect is that I worry the AI functionality will take effort will take development efforts away from improvements to the aspects of Kagi that I do use. It would be nice if it was possible to signal this to the team, even if there is no difference in the price I pay.

          theothertom I understand. We remain first and foremost in the business of search, and AI features are there to enhance that experience for some users like blocking domains and lenses do for others. In other words it a customer acquistion channel for customers who want these features in search (lot of people nowadays do). But our main focus remains on enhancing the core search experience.

            2 years later

            I'm also disappointed that Kagi is paying unethical "AI" companies to spread lies to their users at a heavy environmental cost.

            The only plan without automated disinformation machine support is the basic one. No family plan excludes this pox on society so the messaging from Kagi is clear and unfortunate in this regard, and the only way for customers to message back is to not become one, which is not as clear unfortunately because not many will register on this site and voice their opinion on the subject.

              eobet Kagi doesn’t pay AI companies anything unless you specifically ask AI to be used (Quick Answer is billed as a separate search for this reason). I use and like Assistant and Quick Answer, but am a bit of an AI skeptic so I see where you’re coming from. If you don’t want Kagi to pay for AI on your behalf tho it really is as simple as not using it, AI is opt-in by default and is never invoked without users wanting it to be.

                  eobet I'm sorry to hear that Kagi isn't what you're looking for in a search engine.

                  Best of luck with the other options available. 😃

                      eobet "AI" is just a term for a computing feature that doesn't have common parlance yet. Do you refer to OCR or TTS or spellcheck or algorithmic email spam filtering or pattern recognition or predictive text as "unethical AI" as well?

                      AI is not new. AI being hyped is not new. AI has had several "winters" since the 1970s and before. It's just technology.

                          No one is typing