I don't think you can do anything about this - short of sending quickanswer context to the model i suppose, but that would be cheating. Anyway, I figured I should report it.
When asked about "did steve jobs fire scott forestall", the Kagi Quick Answer tells us that he was in fact fired by Tim Cook.
If one is to click the Agent icon, and switch to chat, chat would tell me the opposite - Steve Jobs actually fired him. Huh?
How, for reference, if we ask Google for it, the quick answer will know that Jobs was already dead by the time Scott was fired (otherwise I doubt Maps would ever be released in the state it was in). And Google is in fact correct, here is a rundown:
- Steve Jobs died on October 5, 2011.
- The Maps stuff happened with the launch of iOS 6 in September 2012 — almost a year after Jobs had passed.
- Before October 2012, Scott refused to sign the public apology about Maps (Tim Cook signed it instead).
- I kinda expected both AIs to tell me the same thing.
- I expected the Agent to actually carry over context from the Quick Answer, thus allowing me to ask further questions - instead of producing an entirely new answer
