I remember Clippy being the face of an online help system for Office that was highly effective. It's hard to remember before Google that the standard of a full-text search facility was that it probably didn't work. People had learned the hard way that it just wasn't worth doing a full-text search for the manual of a software product. Microsoft tried to break through that by putting a better-than-expected search facility behind an engaging character that... didn't work on an emotional level.
Compare to other bad product launches from Microsoft such as OneDrive. Developed shambolically they didn't do due diligence on the trademark and had to rename it from SkyDrive. They tried to force it down your throat when it was in a state where it didn't work reliably: OneDrive became the default save location for Office and if OneDrive was flaking out you could not save a file at all! (Annoying as hell for a platform that is always asking me if I want to recover a file I didn't save four years ago)
There's also the fact that people seriously roll their eyes now when they hear "AI", "Blockchain" and "5G" (especially in the same sentence) There are numerous papers where people apply ML to scientific problems and I'm not so inclined to post them because too many of y'all think that anything like that is a scam... Even though the automatic differentiator in neural network tools is sure helpful for solving differential equations and such.
If I was trying to sell an AI product in 2024 I'd try to avoid the words "AI" and "ML" specifically: if you were working on this stuff ten years ago you were ahead of the curve, if it looks like you just got into it now the automatic assumption is you are behind the curve, not ahead.
I remember Clippy being the face of an online help system for Office that was highly effective. It's hard to remember before Google that the standard of a full-text search facility was that it probably didn't work. People had learned the hard way that it just wasn't worth doing a full-text search for the manual of a software product. Microsoft tried to break through that by putting a better-than-expected search facility behind an engaging character that... didn't work on an emotional level.
Compare to other bad product launches from Microsoft such as OneDrive. Developed shambolically they didn't do due diligence on the trademark and had to rename it from SkyDrive. They tried to force it down your throat when it was in a state where it didn't work reliably: OneDrive became the default save location for Office and if OneDrive was flaking out you could not save a file at all! (Annoying as hell for a platform that is always asking me if I want to recover a file I didn't save four years ago)
There's also the fact that people seriously roll their eyes now when they hear "AI", "Blockchain" and "5G" (especially in the same sentence) There are numerous papers where people apply ML to scientific problems and I'm not so inclined to post them because too many of y'all think that anything like that is a scam... Even though the automatic differentiator in neural network tools is sure helpful for solving differential equations and such.
If I was trying to sell an AI product in 2024 I'd try to avoid the words "AI" and "ML" specifically: if you were working on this stuff ten years ago you were ahead of the curve, if it looks like you just got into it now the automatic assumption is you are behind the curve, not ahead.