As more and more of us run into the limitations of LLMs and other large models, many are starting to wonder whether the technology is overhyped and that it will be more of a fluke rather than turning the world upside down. Also, some, myself included actually, believe that it is likely that we’ll see a bursting of the AI bubble in the same way as the internet bubble burst in 2000. However, even if some large companies need to write off tens if not hundreds of billions of investment in AI due to a bursting bubble, the technology progress has been made and the innovations are available to us all. The internet is a huge industry, despite its bubble being burst 25 years ago, and the same thing will be the case with AI. The only difference is that we will pick up after a bursting AI bubble faster as we have enough people that experienced the internet bubble and will see the parallel. And as I write about in my Sunday blog posts, most companies and professionals are in “playtime” mode, using AI to complement or automate some of their tasks, rather than reinventing entire business processes from the ground up in an AI-first fashion. So, assume that AI is severely underhyped and act accordingly!