ARTIFICIAL INTELLIGENCE: PART TWO

What do humans have that an AI replacement doesn’t? Well, a body for one thing. Okay, so let’s give it one. What does a human have that an AI-driven robot doesn’t? A certain kind of body, namely a human one.

Okay, so how much of my body can be substituted before it ceases to be human? Is there some special organ that’s the without-which-not of being a human? Or even if there is, and it has it, and so it’s human, when does it cease to be me? Is there some special neuron in the brain which is the sine qua non of personal identity?

It might be I have to be human to be the particular me that I am. But can something that’s not human nonetheless be a me? That is, can it have the same kind of thing to make it a me that makes me a me?

I’m using the word ‘me’ here as a kind of abbreviation for what’s sometimes called subjectivity, or consciousness, or awareness, or qualia, or experience, or a mental life … In short, that thing, whatever it is, that we’re pretty sure robots don’t have. Or at least not yet.

Some people think – Descartes famously thought so – that this me-ness has to be implanted, presumably by God. And so far as we know, He only implants it in humans. That, presumably, is what kindergarten Christians think makes the Sixth Day so special. God could implant it in other organisms. And maybe some day He will. But He has to do it. We can’t. We can’t because it’s not made of the kind of stuff other things are made of. It’s sui generis. Even if it’s physical – perhaps it’s made of Mentalese 237 isotope – it can’t be artificed. It had to come from God’s factory.

Theists think God has this me-ness too, though in His case it would have to be Me-ness. In fact it’s hard to think how one could be a theist if he thought God didn’t have a mental life. Which is not to say all mental states are conceptual. I don’t need a concept of pain to be in pain. But I do need a concept of me to talk about my pain. As does God for Him to know what His First Commandment could be referring to. “Thou shall have no other gods before …” Before whom? Oh yeah, “before Me.”

But one doesn’t have to be a theist to believe there’s something magical about consciousness. That it’s not something we can replicate. Other people think we can replicate it, and it’s only a matter of time before we do. But the key question is, How would we know when we’d done it? Whatever the Turing test might be for consciousness, passing it only tells us we can’t tell the difference. But that we can’t tell the difference doesn’t mean there isn’t one. No matter what robots end up doing that we thought only we can do, there’ll be incorrigible holdouts who’ll march through Charlottesville chanting “AI will not replace us!”

Well, let them, say I. No one will accuse you of bigotry if you don’t want to date a black woman. So neither will they if you want to check whether there’s a latch at the back of her head. As they say, chaque a son gout. But the question can’t be ducked if there’s a latch at the back of her head and she’s accused of a criminal offence. The actus reus isn’t at issue. But the mens rea is.

As things have stood until now, how do we establish mens rea? By checking that there isn’t a latch at the back of the accused’s head? Clearly not. So would it be slam-dunk exculpatory if it turns out there is one? Even more clearly not. As Daniel Dennett has convinced us, the assignment of intentionality is dictated by what gives us prediction and control. What’s in the black box is none of our concern.

Ever since Mary Shelley’s Frankenstein – though she didn’t call it this at the time – futurists have been angsting over whether AI will ever achieve consciousness. Let me quote another classic. As Rhett Butler put it, “Frankly, Scarlet, I don’t give a damn!” It’s not that understanding consciousness – what it is and how it comes about – is a fool’s errand. Enquiring minds do want to know. But consciousness has nothing to do with the current angst about AI.

The angst about AI is forensic. And on that score, it’s identical to the angst about theodicy. Who’s answerable for the evil in the world? The proximal agent, be it wetware or hard, or its however-distal Creator?

The lesson of theodicy is that God could have been more careful. So if we’re going to be creators ourselves, maybe that warrants the attention it’s been getting. And then some.



Categories: Angst, Everything You Wanted to Know About What's Going On in the World But Were Afraid to Ask, Social and Political Philosophy

Tags: , , , , , , , , , , , ,

Leave a comment