Sutton's bitter lesson is a crucial point. I see a common criticism of AI from people as "it uses lot of data and compute by applying brute force" and "why don't AI people incorporate human knowledge instead of guzzling so much power" or "it doesn't understand english, its just words statistics" mainly coming from people who are probably not aware of this bitter lesson.
(nit: there seems to be a broken sentence or typo in the last line of the second paragraph in the section where you talk about the bitter lesson)
Btw for RAG you can just use bing copoilet as it can access you document as it’s Microsoft and it has access to your Microsoft file, also you mentioned this in an earlier article sir
Unfortunately, I haven't used Microsoft's Office Copilot (I've used the Bing Copilot and Github Copilot), so can't be sure, but my understanding from other people's experiences is that Office Copilot isn't very good. (Caveat: I could be mistaken; and/or this could change anytime.)
Sutton's bitter lesson is a crucial point. I see a common criticism of AI from people as "it uses lot of data and compute by applying brute force" and "why don't AI people incorporate human knowledge instead of guzzling so much power" or "it doesn't understand english, its just words statistics" mainly coming from people who are probably not aware of this bitter lesson.
(nit: there seems to be a broken sentence or typo in the last line of the second paragraph in the section where you talk about the bitter lesson)
Thanks for pointing this out. Fixed.
Btw for RAG you can just use bing copoilet as it can access you document as it’s Microsoft and it has access to your Microsoft file, also you mentioned this in an earlier article sir
Unfortunately, I haven't used Microsoft's Office Copilot (I've used the Bing Copilot and Github Copilot), so can't be sure, but my understanding from other people's experiences is that Office Copilot isn't very good. (Caveat: I could be mistaken; and/or this could change anytime.)