Heartfelt

So today was the last day of one of my co-workers with our company after nine years. It’s a bittersweet departure. She was an excellent employee—hard-working, super smart, diligent, even-tempered, modest, respectful, and a true team player. I’m not sure of the reasons she decided to leave the company, but it was somewhat unexpected. She was being groomed to become our comptroller.

She is still quite young and unmarried, and I believe she has decided to take some time off for personal reasons before returning to China, where there is a successful family business that she will likely take over. She will be successful in whatever she chooses to pursue in life—there’s little doubt about that.

But this post isn’t really about her.

It’s about the messages sent to her through the company email system to express appreciation and wish her well. We all got to see them. They were beautifully written tributes—accurate in their description of our beloved co-worker, her talents, and her importance to the company. My conservative guess is that 90% of them were either fully written by ChatGPT or, at the very least, heavily edited by it. I made sure mine wasn’t processed through the AI meat grinder.

I know—we can’t all be Shakespeare. And forgive me for sounding like a curmudgeon, but doesn’t it kind of defeat the purpose if you use AI to express something that’s supposed to be heartfelt and personal?

It’s one thing to use AI for marketing, to edit a sales report, or even to help shape a short blog post. I’m guilty of that myself. But this need for perfection—the refined expression, the polished image, the flawless impression—is slowly killing everything.

I miss the sometimes ham-handed expression of genuine feeling. In fact, the one or two messages that clearly weren’t AI-generated were refreshingly obvious by comparison. They were loose and searching, cobbled together and ungrammatical. In other words, they were real.

This small end-of-week episode dovetails with something I heard earlier in the week: Noah Yuval Harari’s talk at the World Economic Forum in Davos. He spoke about what it means to be human. If it’s reason that defines us, then we’re in trouble. "Cogito, ergo sum"—I think, therefore I am—is how René Descartes reasoned the Western human being into existence in the 17th century. Since then, our capacity to think has largely defined our existence.

Harari argues that AI has rendered that definition obsolete. When we build machines that can outthink us in nearly every domain—science, philosophy, academics, mathematics, finance—either humanity has reached a dead end in terms of purpose and meaning, or a new self-definition must emerge.

The alternative seems obvious. It is not thinking that truly defines us. Machines will do that better than we ever could. What truly defines us is feeling: suffering and joy, love and grief, and the expression of those feelings. Machines will undoubtedly learn to fake that—and do it convincingly. But the essential ingredient is still missing. It does not originate from human experience. And without that, any machine-made product is disqualified from being called art.

The advent of AI has suddenly put emotion—and the expression of emotion in art—back at the center of the question of what it means to be human.

Many AI prognosticators may be right that AI could mean our doom. Not because it will send armies to destroy us, but because we may allow ourselves to be infected by it—letting it mutate inside us and quietly alter our sense of what it means to be alive. The only inoculation is to redefine and re-valorize feeling as the essence of humanity.

That sounds like a monumental task in a world where AI is taking over so much of our daily existence. And yet it could begin with something as small as rediscovering the beauty, simplicity, and significance of writing a heartfelt note of appreciation—to a friend, a relative, or a colleague who will be missed.