When I was a student, the general thought was that advances in technology were driven by problems that needed to be solved. President John F. Kennedy delivered a famous speech to Congress in which he set the goal of “landing a man on the moon and returning him safely to the Earth” within a decade. Initially, that goal was not accepted by everyone. Shortly after the speech, 58% of Americans were opposed to the goal according to polls taken at the time. However, the goal was achieved, and the achievement of that goal colored my elementary and high school experience. Achieving that lofty goal was a process of solving a lot of problems, big and small.

I once had the honor of delivering the eulogy at the funeral of a NASA engineer who had headed up a team of engineers that built systems for handling liquid oxygen required for the rockets of NASA’s Mercury, Gemini, and Apollo programs. According to the man’s children, one of his favorite phrases was, “Well, actually, it is rocket science, but that doesn’t mean it is impossible.”

The space program gave us a host of consumer products that we use daily. Some of those products solve problems that come up frequently but for which we had not yet imagined solutions. Velcro and WD40 are two products that are immensely useful that are spin offs of the space program. There are lots of other products that we now take for granted that have their roots in scientists and engineers solving specific problems involved in human space travel.

Sometimes, however, products appear that solve problems that we don’t know that we have. I used to make a comment in that regard about electric windows in cars. “Why do I need electric windows? Turning a crank to lower or raise a widow is a perfectly elegant way to accomplish the task. Having a motor and a mechanism just adds to the complexity and expense of the car. We pay more and gain nothing.” My protests, however, accomplished nothing. Electric windows are here to stay. the last car we owned with manual windows was a 1982 Ford Escort. Ever since, we have driven cars with electric actuators that raise and lower the windows. The devices are reliable and use a small amount of energy and are universally accepted. I still believe that our cars are filled with all kinds of unnecessary technology, but my opinion has no effect on automotive designers.

There are lots of things about computer software that seem to be seeking to solve problems that don’t exist. I am a big fan of computers and I use mine extensively. But some features that come with software updates seem to me to be unnecessarily complex. Initially I resisted spell check programs. The early versions were clumsy and often inaccurate. I used to be so frustrated with the spell check program that couldn’t distinguish between “here” and “hear,” and made mistakes with “to” and “too.” The programs have improved and I’ve learned to use them. The same is true with grammar checkers. Initially, I didn’t want them at all, but I have learned to use a grammar checker to improve my writing.

However, technicians and creators have come up with software that seems to “solve” problems that I don’t think are problems at all. I have several friends who are university professors who are very concerned about the emergence of ChatGPT. About a year ago a free online chatbot appeared that can expertly write almost anything, from English essays to news articles. It can write computer code. All the user has to do is to input a simple prompt. Essentially, the task of writing an essay, for example, is to come up with a good topic sentence. A news article can be written from a headline. The program is enormously popular. Some US universities have had to block it from their servers because of the large amount of students who use to to cheat on their assignments.

It raises the question about artificial intelligence displacing writers and other creative thinkers. Newspapers are already cutting corners by using computers as editors. Is the future one where the news we read is generated without human input? The program’s capacity to write computer code has already been used to create harmful malware.

Reminiscent of the old days when developers were selling radar detectors to consumers and then selling more sophisticated radar to law enforcement in a kind of technology race, creative thinkers are already coming up with artificial intelligence to detect the use of ChatGPT. Edward Tian, a senior at Princeton University has developed an application called GPTZero that has been proven to be highly successful in detecting the difference between text written by a human and that generated by artificial intelligence. The program, now available in a beta version for public use, continues to be refined, but it uses two variables in a text - perplexity and bustiness - and assigns each of those variables a score. AI generated text has little variation in its level of complexity and uses sentences and paragraphs that are very similar in length.

Human generated text rarely is as consistent.

Mr. Tian’s program has been used on easy tasks, such as distinguishing between a BBC article written by human journalists and one written by ChatGPT using the same headline. The program was very accurate with a less than 2% false positive rate. The program has been wildly popular. It has been used by thousands of people and teachers and university admissions officers from around the world have contacted Tian for additional information about how the program works.

The problem that Tian set out to address, academic plagiarism, is a real problem. And it is easy to imagine how the program, or variants of it, could be used to address other real problems, such as detecting computer malware and online disinformation campaigns.

Like electric windows in cars, artificial intelligence is here to stay. Like it or not, we need to learn to live with it. In the meantime, however, I have no use for ChatGPT and have not used it. I continue to write my journal entries the old-fashioned way. Well, mostly - I do use spell check.