This article makes an interesting point about generative AI stuff that I hadn't heard before:
I quietly attended a video conference about possible uses for ChatGPT tech. It was a series of presentations in an open-ended format, little demos of rough drafts and proofs of concept. At random: one to help design a kitchen. It was cool. You could whip up a little kitchen and do a little walkthrough and make changes. Would the kitchen collapse if actually built? Probably not. Did the AI understand the concept of load-bearing walls? Unknown. Did it know what it would be like to actually work in the kitchen it built? 100 percent no. Would somebody trust it? Probably, unless they knew otherwise. Would they pay an expert who had these answers some hundreds of dollars instead of using a nearly-free AI tool? No chance. They might take the AI’s output to an actual expert who could correct things and know the right questions to ask, but the damage has been done, and a whole field begins to suffer, as the demand for entry-level expertise vanishes, gutting the growth of a profession at the root.
I don't know to what extent to expect this, but I can absolutely see that without the need to hire people to design cheap little icons or brief copy text, lots of industries might find themselves with a dearth of trained talent because no one is even trying to do that anymore, resulting in an overall quality decrease because there's less practice going into the art overall.
I wonder if there are comparable occurrences in history, but those are all pretty grounded in like, manufacturing, which seems different in terms of the tradeoffs you're making. There's probably a textiles-related thing here if we want to wrap around on the Luddites.
















