I do not see us reach AGI within the next ten years at last, probably in the next 20. Deep Learning and LLMs have taken us down a road that provides many such cheat codes, but is not AGI, and is ultimately is self-limiting in getting to producing an AGI.

I can see more people adopting a neo-amish style approach though as LLMs become more prevalent.

The problem I also see is that as the backlash increases against LLMs in specific, and AI in general, the marketers will mention AI in their product descriptions and web sites less and less. I think it will get to the point where you don't know whether a product uses LLMs or not. Alternatively, it will get to the point where LLMs are ubiquitous, and you can't easily find products that don't use them.

As a result, I think they are a part of life now. We can't ignore them, legislate them away, or decide not to use them without major impact on our lives. Instead, we need to legislate responsible use, legislate attribution and compensation, and learn to recognize the signs of GAN generated content.