The issue there is that even at that pricepoint, Microsoft is still operating CoPilot at a loss. If they drop it more, they’ll be making even more of a loss. Which is the standard business model for new products these days, but the losses on AI products dwarf things like Netflix and Uber during their “operate at a loss to drive everybody else out of business” phase.
Of course, that would all be fine if CoPilot was some killer product that people quickly found themselves unable to work without. Instead, the feedback shows that workers find that it’s not useful or reliable enough to be worth using, and Microsoft’s own latest advert for CoPilot in Excel contains data which shows that at best operation it doesn’t work 46% of the time, and that figure can be as high as 80%.
I’m not sure these problems are really surmountable - you’ve got an incredibly expensive-to-run product which doesn’t do much that’s useful and is bad at the things that it actually could be useful for. It’s not just Microsoft, it’s the entire tech industry that’s facing this problem.
echodot@feddit.uk 3 weeks ago
This is already happening.
I work for very large IT company and they are upgrading to Windows 11 because they have to but AI tools like co-pilot are being blocked by default in the image we push to all users.
This is resulted in a very funny knowledge base article which basically tells the support staff to tell the users to go do one if they complain about it.
shalafi@lemmy.world 2 weeks ago
Any job openings for a sysadmin?
Evotech@lemmy.world 2 weeks ago
Banning work controlled ai is incredibly short sighted
You will just end up with a whole bunch of shadow ai that you can’t control what users upload or how they use it.
echodot@feddit.uk 2 weeks ago
Well co-pilot sends all it’s data to Microsoft as is.
Broken@lemmy.ml 2 weeks ago
Is that an internal KB article or something you send to the customers? If it’s public I’d like to read it for a chuckle.
echodot@feddit.uk 2 weeks ago
Customers in this case being other businesses, so no it’s not public.
Basically it just boils down to telling them that corporate policy is not to make use of AI because of data collection concerns and to emphatically remind them that attempting to circumvent these policies is against corporate procedures.
Technically it’s no different than just taking internal documents and dumping them online.