I’m way older than Kyle Hill’s target audience, but I can still enjoy watching him. I’m not the only one: As a YouTube science educator he’s pushing 2 million subscribers, and he’s also a White House science advisor, his bio says.

Where Hill is good, he’s very good. Unfortunately like a lot of other science educators and journalists — and even scientists — he can turn dangerously misleading at times.

In the case of artificial intelligence it comes down to one sentence: “This is going to change our relationship with each other, with work, with technology, with society as a whole, because human intelligence is no longer exclusively human.”

Here’s the context of that sentence, delivered in a short 2 1/2 minute excerpt from a video he streamed on the rise of artificial intelligence. AI is both exciting and terrifying, Hill says, which is true enough, except he should be also terrorized by the mistake he himself makes over it. It’s not just him: I see the same mistake everywhere, so we need to know about it and be prepared to answer it. This excerpt provides a great example, because it’s so short yet speaks it so well.

(The clip is set for specific start and stop times. Elsewhere there’s a commercial break with content some readers may find offensive. You can push past this excerpt, though not without intentionally trying. Use discretion if you do.)

AI: There’s No Stopping It Now

He’s absolutely right: AI is coming, and there’s no holding it back. “If the United States stops AI’s development, China is not going to.” It is going to change the world, and it’s happening fast.

Jump ahead to 11:40 (that’s a safe place) and you’ll hear him quoting “misinformation expert” Nina Schick predicting that within two years, 90% of the “information” you find on the internet will be synthetic — not necessarily false, but not generated by humans, either.

For that reason and others, he warns us,

You have a front-row seat to a very interesting next few years. It feels to me like another one of those great upheavals in perspective is coming, like when we found out that we weren’t the center of the universe, that earth goes around the sun, that we weren’t the pinnacle of evolution. You have a front-row seat to another species-wide realization: That the last thing that we thought made us so uniquely human no longer makes us special.

Yuvale Noah Harari sees something like that in our future, and says it’s about some kind of next stage in evolution. He’s even said we could unite the world’s religions by having the world’s first “superhuman intelligence” write a “holy book” everyone could agree on.

AI’s Greatest Dangers?

There’s a lot to be worried over here. AI promises new frontiers of knowledge and productivity. With that we have the widely known problems of AI-generated false “information,” deep fakes, and the huge economic and social shake-ups barreling down on us all.

Everyone knows that, everyone agrees on that.

Worse Than That

Unfortunately, not everyone sees the mistake that’s going to cost us even more in the long run — the mistake Hill and so many others make over and over again. Losing your job to a machine would be awful, but it would also be nothing terribly new in human history. Giving up what makes us human, what makes us special, though? That points toward an end to human history. If human intelligence is no longer the unique thing it’s always been, if a “superhuman intelligence” could soon rewrite religions’ holy books, what would it then mean to be human? What happens to human dignity when we can all click into a chat program that’s smarter than us all?

Mark my word: This will cause more damage than anything else you’ve been warned about.

The world’s greatest terrors have always come from humans denying others’ humanness. In the past it’s been selective: The Nazis’ reputation for the 20th century’s greatest evil doesn’t come from the numbers of people they killed. Stalin and Mao both outdid them by millions. It comes from their assembly-line methods, and their using Holocaust victims’ hair as raw material for manufacturing. They treated humans as non-humans.

Likewise, though on a different level, I take it that the Civil Rights movement in the 20th century was misnamed: It wasn’t a movement for civil rights, it was for civil recognition of blacks’ full humanity, allowing them full participation in what American had always considered human rights.

It Doesn’t Even Take a Tyrant

And I think it will remain selective, but here is where the true terror awaits: The more people who call AI a superior form of intelligence, the more the entire human race will be selected for this kind of dehumanization. AI won’t actually rule. It doesn’t need to, if we think it does.

That’s a reminder of what can happen when powerful people or institutions start dehumanizing others. It doesn’t take a tyrant to ruin a life, though. When humans are presented as having no special meaning or purpose, you can expect individuals will conclude they have no special meaning or purpose. How much more depression, anxiety, and suicide do we need before we recognize our reigning ideologies are pointing people in exactly that direction?

When humans are regarded as less than machines, you can expect them to become as disposable as machines. Assisted suicide, already skyrocketing in Canada, will soon be one of many ways to eliminate that which is old, worn out, or in the way. Undoubtedly “in the way” will include, “in the way of our political progress.”

Reality Check: This is Intelligence?

The good news is, there’s no truth to the rumors. AI isn’t more intelligent than humans. In fact it knows so little, I’m not sure the word “stupid” even covers it.

Take ChatGPT, for example. It produces some impressive writing, no doubt about that. But do you know how it does it? Exactly one word at a time. The way ChatGPT produces a sentence, a paragraph, a paper or a thesis is by examining its database and predicting the next word, one word at a time. It does it almost entirely by probabilities: Based on the gazillion words in its memory, what’s the most likely word to follow after … ?

I’ll leave that blank. You can pretend you’re ChatGPT. You might say the most likely word would by “this.” Then you’d start over again: “Based on the gazillion words in its memory, what’s the most likely word to follow after this … “? And you might say “one” is the most probable next word.

That’s a little bit like AI, but not close enough. You’re not predicting based on a databank of words you’ve experienced being strung together. You’re predicting based on what words mean to you. AI researchers will talk about “recursion,” “context,” “learning,” and even “understanding,” but it’s still the case that it works a word at a time. And then it starts over, re-examines everything as if from scratch, and produces the next one.

You, on the other hand, couldn’t do it one word at a time if you tried. Humans construct sentences based on complete thoughts. The “tip of the tongue” phenomenon shows it’s possible to construct thoughts even without knowing the next word.

Seriously?

Meanwhile, even if it were true that ChatGPT could do something like what we call thinking, especially producing new thoughts. Those thoughts would we limited to exactly one word at a time. It uses strict mathematical algorithms, which would have the boring result of producing exactly the same output every time if not for a touch of humanness the programmers add. That “humanness” comes from randomizing a percentage of its operation. Randomness is the ultimate in non-intelligent. Does anyone think the dice are thinking about what numbers they should roll to? And that’s for added “humanness”?

And to wrap it up, when’s the last time you heard anyone touting artificial wisdom?‘

Nothing to Fear But …

AI is no threat to humanity. It’s never going to outsmart us. It’s a counterfeit. Impressive in many ways, but still counterfeit.

What I’m concerned about instead is the people developing it, controlling it, and unleashing it. It’s going to disrupt almost everything, and we’re not prepared for that. I don’t know that we ever could be. One good step in that direction, though, would be putting an end to all this nonsense about a machine having intelligence.

We may be too far down the road to stop it even that small amount. Regardless, you can stand your own ground, and you can teach your family, your church, anyone who will listen: Humans are still as unique as ever. We’re still the only creature made in the image of God. We have no competitors to that title — or to the dignity and worth that come with it.

 

Tom Gilson (@TomGilsonAuthor) is a senior editor with The Stream and the author or editor of six books, including the highly acclaimed Too Good To Be False: How Jesus’ Incomparable Character Reveals His Reality.