The world is atwitter about Artificial Intelligence these days. As they should perhaps be. 

Humans have an innate ability to do two things well. We can recognize patterns and learn from past experiences, which can in turn inform our realities and interactions with the outside world. And, we can synthesize ideas of ourselves and others to create new ideas or apply old ideas to new realms. 

We must also perform more mundane intellectual tasks that merely record and classify events around us so they can be documented for others and posterity. Once in a while, one can even create brand new knowledge, but that is relatively rare and resides in the forefront of science, the arts and humanities, and original research. Beyond these creative acts that are completely original, these other intellectual exercises are mundane, but necessary, and perhaps even comforting and rewarding. 

We all wish we had the patience and time to most thoroughly inform and document our every observation and work product. For many, such thorough research is tedious, time consuming, and even exhausting. 

Machines know no such exhaustion, though. The World War II decrypter and early computing pioneer Alan Turing stated that someday the tireless energy and knowledge storage capacity of computers will reach a point that they can answer a question from a human in a way that humans can no longer determine whether they are speaking to a computer. The state of AI has now passed that Turing Test. AI takes advantage of Moore's Law that states computer power doubles about every 18 months. AI has now gone exponential, as today's graph demonstrates.

This milestone should be liberating. Think of the activities that AI now frees us from. They include basic legal analysis and documentation, document editing and even the creation of passable essays, documentation of medical records, the reading of test results, and the comparison of symptoms to a vast array of potential causes, personal coaching, the teaching of courses, analysis of resumes, and, of course, the generation of search results in our browsers or computer-based personal assistants. 

All these functions are performed now and far more often than any of us would like to believe. They can leverage the skills of a doctor, lawyer, feature writer, or consultant two or three times over as these professionals are freed from the most mundane parts of their jobs. 

AI can generate art, compose music, and create simple computer programs. It can recognize faces in crowds far faster than could a roomful of humans, and can navigate a vehicle at least as well as most drivers. And it is just getting warmed up. 

About a decade ago China developed an industrial policy that left chip design and scientific programming to other nations. China instead is determined to corner the world in AI and in the hardware necessary to support its power that is doubling in a matter of years rather than decades. If there is something that AI can do passably well now, remember that it could not perform those functions at all just a few years ago, and will likely perform them better than many humans a few years hence. 

Robotics and automation are making much manual labor unnecessary. AI is making a fair amount of intellectual labor obsolete too. These developments will make billions for the robot makers, the automated manufacturers, and the IBMs, Googles, Microsofts, Apples, ChatGPT, and the Chinese that are rapidly improving the machines that could make obsolete about 30% of the workforce in just a couple more decades. 

With increased robotization, it was always argued that they will make those employed in manufacturing more efficient. We now have about 50 years of increased automation behind us, and indeed each worker is more efficient. But, the number of workers in manufacturing has declined precipitously. Instead, those no longer employed in manufacturing were smart if they learned to build, wire, and plumb houses, or become service workers. Automation will improve home construction, but we will always need plumbers. We need relatively few people in agriculture and manufacturing anymore, and fewer all the time as AI takes the tasks away that we described earlier. We have also reached a point where AI can code better AI. Our machines can build better machines, with little or no human intervention.

At the macroeconomic level, there will be a lot of production, but with relatively few people earning the profits of production. Without the billions of workers in the world who are out-manufactured by machines and out-thought by AI, it is hard to imagine how people will gain a livelihood to purchase all the production that will only become increasingly inexpensive. Ultimately, those who will find employment have such skills that machines cannot replicate, such as the ability to dunk a basketball, or offer their skills at such a low wage that AI won’t bother for the trifling profits. There will always be those few that produce something so local, such as a massage or a tooth extraction, that the intellectual and diagnostic prowess of AI can’t replace. 

My lingering question is whether AI is up to the task to figure out how humans can garner an income and maintain our dignity when so many of our efforts are so easy to reproduce digitally. Maybe AI will become so satient that it realizes that’s a problem they have no interest in solving.

By the way, did I actually write this article? Only AI knows for sure.