As a general thing, I don't think art will "die". That most of the market need, at some point, can be satisfied without human intervention, will never prevent anyone from designing something. It simply won't be a mass commercial endeavour (or maybe a PWYW model), but it could serve a niche market (much like there are still painters who paint landscapes in front of monuments to sell to tourists, despite most of them being satisfied with taking pictures on their smartphones), or they could just produce art for art's sake, without making a living from it, as a hobby. It can reduce the number of actor, but "dying" seems too extreme to consider. Also, there might be companies (like Enworld) that might bear the cost of using human-produced art for their customers, and catering to a market of consumers of human-produced art, like there are companies specializing in organic vegetables, despite most of the agricultural sector adopting industrial fertilizers.
Whether I'd buy a 100% ai-made product by WTC? I wonder. I'd have to determine whether the AI made by WotC is worth paying the price over what I could obtain by either running an LLM at home or what will be available through subscription models. There is a strong chance WotC won't be able to rival with the state of the art product, and the help an AI-designed product by their prompters wouldn't help me more than directly asking an LLM to assist me in my gamemastering task. So I most probably wouldn't buy a WotC made, 100% designed by AI, but not on the basis that it's made by AI but because there is a strong chance it would be published at a time where AI assistants are good enough to compete with the product they could sell. Unless they sell it really, really cheap.