My local authority uses Microsoft Copilot 365. It used this recently to respond to a complaint I made regarding more than one serious failing by their Adult Social Services teams.
Co-pilot summarised my complaint, produced a response to it, and then quite touchingly included a link to a local authority web page that cautions the public against the use of AI to write complaints to the local authority, possibly on the “assumption” that my complaint was itself largely or partly the product of AI, when in fact it was neither.
The summary wasn’t atrocious, but it saw no hierarchy of emphasis in the different strands to my complaint when an intelligent human reader would have identified this hierarchy instantly. Identifying it was crucial to a sound understanding of my complaint and the rectification I was seeking.
If human beings employed in language processing roles are no longer required to process language using their own brains, then after a while their ability to do so without the aid of AI will almost certainly be less good than it was to begin with.
One of the beauties of AI in my view is that, counterintuitively perhaps, far from levelling the playing field between people with different levels of intelligence, it creates even greater disparity, inasmuch as an intelligent person using AI intelligently can arguably outsmart a less intelligent person using AI even more easily than if AI was not part of the mix.
Ultimately, at every level of its existence, AI is only as good as the humans behind it.