Kagi is an ad-free, premium search engine for people who value their time, attention and privacy. With results optimized to serve you and not the advertisers, Kagi is lightning fast and offers advanced features for customizing your search experience.
Basically the title.
I'm interested in any opportunity to inprove the way I navigate the internet. What I've been for a few years now is DDG, which works fine. Not great, not amazing, just fine. And that's ok considering how they opperate.
I just heard about kagi and was really cosidering it. Makes sense as a business model (pay so we don't have to sell you data), seems privacy respecting, and claims to strive for best search results in the market. Some test searches from the trial seem promising.
If you've used it for any amount of time, what has your experience been with it? What plan are you using? What are you mostly searching for?
Even you haven't used it, any thoughts / opinions are welcome.
Haven't used it, but according to their Privacy & Terms they started using FastGPT which is a dealbreaker for me. I'll stick to SearX which allows more curation.
they started using FastGPT which is a dealbreaker for me.
Care to elaborate on why? I haven't being keeping up with all the AIs, and a 3 second search isn't returning anything nefarious.
Accuracy. They are known to hallucinate. Sifting through various sources to verify information is already time consuming task without AI created nonsense that is impossible to source check.
FWIW, the AI features are not used to provide search results; they are all on-demand and triggered by the user (via Quick Answer, or Universal Summarizer, or the "discuss this site" feature).
The founder is well aware of the problems with AI and that is taken into account when deciding how to use it in Kagi.
Generative AI is a hot topic, but the technology still has flaws. Critics of AI warn that “[AI] will degrade our science and debase our ethics by incorporating into our technology a fundamentally flawed conception of language and knowledge”.
From an information retrieval point of view, relevant to our context of a search engine, we should acknowledge the two main limitations of the current generation of AI.
Large language models (LLMs) should not be blindly trusted to provide factual information accurately. They have a significant risk of generating incorrect information or fabricating details (confabulating). This can easily mislead people who are not approaching LLMs pragmatically. (This is a product of auto-regressive nature of these models where the output is predicted one token at a time, and once it strays away from the “correct” path, for which the probablity grows exponentially with the length of the output, it is “doomed” to the end of output, without the ability to plan ahead or correct itself).
LLMs are not intelligent in the human sense. They have no understanding of the actual physical world. They do not have their own genuine opinions, emotions, or sense of self. We must avoid attributing human-like qualities to these systems or thinking of them as having human-level abilities. They are limited AI technologies. (In a way, they are similar to how a wheel can get us from point A to point B, sometimes much more efficiently than human body can, but it lacks the ability to plan and the agility of human body to get us everywhere a human body can)
These limitations required us to pause and reflect on the impact on search experience, before incoporating this new technology for our customers. As a result, we came up with an AI integration philosophy that is guided by these principles:
AI should be used in closed, defined context relevant to search (don’t make a therapist inside the search engine, for example)
AI should be used to enhance the search experience, not to create it or replace it (similar to how we use JavaScript in Kagi, where search still works perfectly fine when JS is disabled in the browser)
AI should be used to the extent that it enhances our humanity, not diminish it (AI should be used to support users, not replace them)
That is better than most other cases, but far from perfect. It can still be wrong and that's even more harmul in "Quick Answer, or Universal Summarizer" as people are more liekly to trust it's result instead of double checking with another source.
That's completely understandable.
In my experience when I have tested the summarizer it has done well at summarizing only what is there. It also splits it up and cites the sources.
Most people likely won't look at the sources, and you are right that is putting a lot of trust in it, especially if they don't understand how the tech works.
I find it good for saving time on something I just need a quick answer on to solve an immediate problem at hand.
For example the other day I asked it for the rules to Stratego. It listed it all right there and pointed out where there was any disagreement between various versions of the game. The stakes were low if something was wrong, and I saved several minutes trying to piece together and remember all the rules from various sites.