Thanks, AI, but I Got This One

This week, we return to AI, this time not to watch Poe grappling with the mystery of the mechanical Turk, but rather to consider the role of AI in our own time.

Have you ever asked ChatGPT or another form of generative AI to write a report for you?  It's an appealing idea, especially if you don't particularly enjoy writing reports.  The notion, however, prompts a funny thought to me.  Is the next step to have AI read that report?  If so, what's the point?

You can read that last question as a rhetorical, even snarky one, but let's take it seriously.  Too often — way too often, if you ask me — we do things without seriously considering the point.  In my work as an academic administrator, I have been involved in creating and reading more reports than I ever would have chosen myself.  I like both writing and reading, but writing and reading reports for me must be like doing and watching the Macarena to a professional ballerina.  It's mind-numbing, almost painful.  Why wouldn't I want to farm this work out to AI?

Well, maybe I would, but the first question is whether we need those reports at all.  What's the point?  In a letter, Henry David Thoreau once wrote, "It is not enough to be industrious; so are the ants.  What are you industrious about?"

When we have answered this question about any particular task, we then can take up another important question, the one that a lot of people are asking these days: What role should AI play in all of this work?

Who gets to think?

Last week, I shared the story of Edgar Allan Poe's effort to solve the mystery of the mechanical Turk, which supposedly could defeat human beings (including Benjamin Franklin and Napoleon) in chess.  In his essay about the Turk, Poe distinguished between the kind of "thinking" that Charles Babbage's early computer could do and the kind required to play chess.  The latter must be far more robust because it involves decisions among various alternatives rather than a purely linear process with no "branches."

Poe naturally would have been one to think in this way about thinking because Poe liked to think.  So do I. 

Because I like to think, I am not quite ready to farm out all of my thinking to AI.  I'm willing to admit that AI already is better at some kinds of thinking than I am and even that it might eventually be better at all kinds of thinking than I am or any other human is.  When it comes to devising medical treatments for deadly diseases, designing better buildings and bridges, combatting climate change, and improving our world and lives in a thousand other ways, I certainly will not be the one to get in the way of progress.  Letting my ego take precedence over progress would be not only foolish, but also selfish.  One of my core values involves service to my fellow human beings.  Insisting that humans do things that AI can do better runs counter to that value, so . . .

Have at it, AI.  Use your smarts to make a better world.

I'm not ready to stop thinking just because we have AI, though.  That strikes me as equivalent to having AI watch a baseball game for me.  Where's the fun in that?  Are we going to have AI do our crossword puzzles, play our board games, go on dates for us?

Recreational Thinking

I suppose it's no surprise that I like to think.  That's why I became a professor, and it's why I enjoy doing research and writing, but I don't think I'm all that unusual.  Using our brains to consider, question, admire, investigate, debate, and create is at the core of the human experience. 

If you need evidence for this assertion, consider all the thinking that you and probably everyone you know do just for fun.  Long before AI became really good at doing some of our thinking for us, we already had countless fellow humans who could do the same thing — that is, think better than we could about a variety of subjects, but virtually all of us still routinely insist on doing our own thinking about politics, sports, and more, often fancying that we were outsmarting the experts.  Why do we bother?

Let's take an innocuous example.  I love baseball, and, having played and studied it for years, I know a lot about it.  Now, I also realize that professional managers know even more, but I never hesitate to question their judgment even while I know very well that they not only understand far more than I do, but also have at their disposal data analytics that I don't have.  "This pitcher is done.  He can't throw the ball over the plate!  Take him out.  TAKE HIM OUT!"

Why do I bother?

Maybe I just like to think — in this case about baseball, but also about books and words and art and thought itself.  Thinking is stimulating.  It's fun.  It makes me feel alive — even if I'm wrong.  (For the record, sometimes I just might be right about taking that pitcher out.  I mean, look, he can't throw the ball over the plate!)

Now maybe you don't like to think about baseball at all.  Maybe you would rather think about birds or music or business.  The point is you like to think.  That's because you're a human being.

This is where all of us — both individually and collectively — should be considering the role that AI will play in our lives.  It can do some of our thinking for us, but we still get to decide where to deploy it.  The question is, to paraphrase Thoreau, "What are we thinking about?"