COLUMBIA, MARYLAND - One cannot escape discussions about artificial intelligence (AI) these days. Not only do I continually have such conversations with colleagues, clients, and content creators nearly every day, but I also have persistent engagements with neighbors, friends, and family concerning AI. These exchanges frequently start with questions like this: Is AI safe? Is AI becoming smarter than us? Is AI destroying jobs? Is AI displacing workers? Is AI out of control?

With explosive new developments this year in the realm of conversational AI chatbots, large language models (LLMs), and ChatGPT - the new ‘big kid’ in the AI world- the globe is buzzing with questions like the ones above. One of the developers of ChatGPT recently said the following: “AI could surpass humanity within the next 10 years as 'superintelligence' starts to exceed other powerful technologies.”¹

So, the question today is: ‘What happens when AI outsmarts humans?’ This article will begin with a story that establishes the context in which to answer that question.

My career background is in astronomy and astrophysics, and my doctoral dissertation was in this field as well. For the first 20-plus years of my career, I was a research scientist and project manager in the United States National Aeronautics and Space Administration supporting data systems for space science missions. The focus was primarily on data systems, data management, data analysis, and data distribution of space science datasets to the worldwide research community. Then, one day, our data center was offered a 2-Terabyte dataset for management and distribution to the world. That might not sound like a big deal by today’s data volume standards, but in 1998 two terabytes was an enormous data volume challenge for a data center. In fact, that one single dataset was two times larger in volume than the combined aggregate total volume of all the other 15,000 datasets in our cente