What are the chances that an AI computer model breaks how trading works in the next few years?

ChatGPT being one obvious example of the many AI models in development- How do the MM's shift when the average joe starts improving based on nothing more than AI becoming more ubiquitous? I swing by following somewhat predictable MM algorithms. If dumb dumb me can do it then I can't imagine general AI models capable of interactive discussion are far behind at this point.