ChatGPT, Conspiracies, and AI-Human Relations

An edgy rap about my conspiracy skepticism and what it means for human motivation and meaning

A friend of mine named Jordan Stern tasked the new AI technology ChatGPT to “write me an edgy rap about Michael Shermer debunking conspiracy theories.” Here is the result. Although I've been skeptical of claims of imminent AI sentience, I must say this is pretty impressive.

Verse 1:

Listen up, it's time to get real,

Shermer's in the building, time to feel

The heat of skepticism, no room for doubt,

He's got facts, and he's shouting it out.

Chorus:

Debunking all the theories, all the lies,

He's the skeptic king, with open eyes,

No stone unturned, no case unsolved,

Shermer's on a mission, to resolve.

Verse 2:

From aliens to hoaxes, he's seen it all,

But he don't fall for it, he stands tall,

He uses logic, science, and reason,

To expose the truth, in all seasons.

Chorus:

Debunking all the theories, all the lies,

He's the skeptic king, with open eyes,

No stone unturned, no case unsolved,

Shermer's on a mission, to resolve.

Verse 3:

So when you hear a conspiracy, don't you panic,

Just call on Shermer, he'll be fantastic,

He'll break it down, with wit and grace,

And leave you with the truth, in its place.

Chorus:

Debunking all the theories, all the lies,

He's the skeptic king, with open eyes,

No stone unturned, no case unsolved,

Shermer's on a mission, to resolve.

Outro:

So don't believe the hype, don't get fooled,

Shermer's got the skills, to keep it cool,

He's the master of skepticism, no contest,

And he's always ready, to do his best.

Another friend, Michael Aisner, who forwarded me Jordan’s ChatGPT query, followed it up with this ominous commentary of what he would task the AI to do:

Dear ChatGPT -

Please write Michael’s next book of all original material based on what he’s been thinking about, so he can go ride his bike in the Andes.  

It’s coming Shermer!!

Whatever the “it” is that is coming, it is surely not sentience, but it is impressive.

(In any case how would we know if an AI is sentient, inasmuch as it could be programmed to say and do things that appear sentient even if inside the hardware there is no self-awareness, no consciousness. This is the problem of “other minds” that remains an intractable one—how do I know that you are sentient? What if everyone on Earth is a “philosophical zombie” just lumbering around in sentient darkness only appearing to be consciously aware? My solution is to apply the Copernican Principle to myself: I’m not special. If your brain is wired up similar to mine, and I am sentient and self-aware, then very probably—let’s put it at 99.999% probable—so are you. As for determining whether or not Data on Star Trek is sentient—or whatever comes after ChatGPT—I am open to suggestions as I don’t know.)

Given that all of the assignments for my Skepticism 101 course at Chapman University are written essays—short-essay quiz questions based on the lectures and long-essay midterm and final questions based on the readings—I have to seriously consider the possibility that students could turn in ChatGPT-written answers and there may be no way for me to know.

I have seen online claims that there are programs you can run on written content to determine if it was produced by ChatGPT, but do I really want to invest the time to do that for 125 essays (25 students x 5 essays) for each of 5 quizzes, and 250 essays (25 students x 10 essays) for the midterm and final each? No, I don’t. As I told my students when one of them asked about ChatGPT (recalling from memory here):

If you’re going to have ChatGPT write all your course essays so you don’t have to read the books or listen to the lectures, then what’s the point of going to college? Why invest four years of your life and a quarter million dollars in tuition and housing costs, when you could skip all that, go out into the world now, and create a great life for yourself? I know many people who never went to college who make good money and lead healthy, happy, and fulfilled lives.

To this I would add that, along these lines, I could also pay someone to ride my bike for me so I don’t have to go out and suffer and sweat in trying to keep up with my fellow cyclists, or (if the AI optimists are right) program a robot to raise my children and sexually service my wife, but what would be the point? I like riding my bike (along with raising my children and being intimate with my wife)! Ditto Aisner’s comment above that I could hire ChatGPT to write my next book. But what’s the point? I like writing books!

Whatever is coming, use it as a tool only and not a replacement of your mind, which must be actively engaged in the world to be of any use.

Michael Shermer is the Publisher of Skeptic magazine, the host of The Michael Shermer Show, and a Presidential Fellow at Chapman University. His many books include Why People Believe Weird Things, The Science of Good and Evil, The Believing Brain, The Moral Arc, and Heavens on Earth. His new book is Conspiracy: Why the Rational Believe the Irrational.