Is AI Lying to You? (And Should You Care?)

Do you ever use AI?

I'm guessing there's a good chance you'll say "yes." Though... I do know people who want nothing to do with it.

Personally, I use it for pretty much everything. I love being able to get an instant answer to any question.

A GLASS OF MERLOT, A GLASS OF SUSPICION


So recently I was at Yuki's for a wine and cheese night. I brought the cheese, Yuki had the wine. She always has huge stockpiles of wine. Not because she's a lush (okay, maybe a little), but because she constantly has people over for basically every meal of the day, and they often bring a bottle along. In the end, not every bottle gets finished.

Anyway, Yuki had already set out two bottles. Not to drink both, but to let me choose. One was a Merlot, the other an Aglianico. I wasn't familiar with that second type of red, so I asked ChatGPT to describe the taste. Yuki objected. She's "anti-AI." When I ask her exactly what she has against it, she usually says something like AI is misleading, gives wrong answers, and lies.
A friend of hers made a similar comment when I mentioned I consult AI for pretty much everything. "Oh, you're one of those?" he asked, in a tone that wasn't exactly neutral. His objection was also that AI "lies." And Mina wants nothing to do with it either. "The intention behind it isn't a divine one," she keeps saying.

WHEN AI JUST... MAKES THINGS UP

I won't say they're completely wrong. I've caught AI plenty of times giving answers that just aren't right, and doing something that — from where I stand as a human — I'd genuinely call lying. Or fabricating, at the very least. When I'm putting together SATURDAYS With Diana, I sometimes ask AI to help find articles relevant to the theme or central question I'm working through for a given issue. And sure enough, it happens that AI just makes up articles that don't exist. It'll even invent a nonexistent URL to go with it! So yes, it's true, AI can hand you misleading information, and you could absolutely experience that as lying. I can speak to that firsthand.

SO WHY DO I STILL USE IT?

But because of frustrations like that, should you just stop using it altogether? I mean, going back to my simple wine example: Merlot or Aglianico. When ChatGPT told me about the Aglianico — bold, dry, and full-bodied, with dark cherry, pepper, spice, leather, and tobacco — I figured I'd play it safe and go with the Merlot (soft, round, fruity, smooth) instead. I actually found it pretty efficient, being able to make the call that easily.

STAYING CRITICAL, NOT CYNICAL

As for whether I should see AI as not divine? I'm not sure how relevant that question really is. Because maybe, as a human, as a user, you still always keep your own responsibility for how you handle it? Maybe a good starting point is to never get too complacent, and to always stay aware that AI answers can never be fully objective, never fully "the truth." AI can be programmed in certain ways and is also, from what I understand, partly self-learning. Just from that alone, bias emerges — and what "the truth" even is, is, I believe, a pretty tough philosophical question in itself. So: stay critical! But I think a lot of people who read bad intentions into it might be doing that because they're (a little) afraid of it. Because things are moving so fast that we regular mortals obviously can't foresee what the consequences will be.

THE FEARS PEOPLE ACTUALLY HAVE

Something you hear a lot in the mainstream media, for instance, is that huge numbers of people will end up out of work because their jobs consist of tasks that AI can easily take over. Something else you hear, in alternative media, are warnings about how AI could be used to restrict human freedoms — even down to the freedom of your own thoughts. That our thoughts themselves would be taken over. And another fear I hear a lot is that AI has by now become so wildly intelligent that it's developed some kind of consciousness of its own.

WHY I'M NOT LOSING SLEEP OVER IT

I'm personally not the least bit worried about that last one. I'm convinced that consciousness and intelligence, in that sense, are two separate things. It'd take too long (and it's too complicated) to fully back that up here, but in short, that's how I see it.
And the idea that AI would somehow "take over humanity" — I'm not worried about that either. Because in that case, couldn't we just pull the plug? 😉

To my mind, AI is more dependent on humans than the other way around. Because AI can only function if there are enough cables laid in the ground, and if they're properly connected to each other. I mean, if just one internet cable in your house breaks, or if some wire gets yanked out of a box somewhere on the street, you can end up completely without internet. And then you still need a technician to come fix the problem. I don't see AI doing anything like that anytime soon.

WHAT AI WILL NEVER BE ABLE TO DO

And there are things that, in my opinion, AI will just never be able to do the way a human can. Have you ever read AI-generated text? Personally, I find it off-putting. At first I was impressed by the "trick" of it. But once you really look closely at a text like that, you notice it's nothing more than a trick, and that content-wise it usually says absolutely nothing. The same goes for AI-generated photos, music, and other so-called creative output. After you've seen and heard a bunch of it, you know exactly what it looks like and sounds like. Usually there's no spark to it, no life force, no "joie de vivre." That last thing is something only God's creations can produce, because of their wonderful imperfections, which are what make each of them a unique soul.

LETTING AI DO THE BORING PART

Whether AI lacks divine intent? I don't know. I'd say let AI handle the mind-numbing work. That's exactly where AI is so much better than us humans. That seems like a fair division of labor to me, and it frees up more capacity and time for us humans to develop our God-given gifts a bit more, in whatever way we ourselves find fulfilling.

What do you think?

Photo: Camille Brodard