Artificial Intelligence - A Convenient Scapegoat

We have a flawed relationship with credit. Not the financial kind, the esteem kind. When something goes well, we tend to give most of the credit to the person we last interacted with (also known as last touch attribution bias¹). When our doctor helps our cancer go into remission, they get the lion's share² of the credit. Not many people write letters to the rest of the oncology team. Far fewer write a note to the pharmaceutical company that made the drug that put their cancer into remission. And almost no one writes to appreciate the people who administered, financed, and ran the clinical trials that made the drugs possible. I don't say this judgmentally - it's a part of the human experience. We appreciate what we can sense directly.

Blame is often treated the same way. The player with the game on the line makes a mistake, that's on them. The server brings the wrong order, it's their fault. Even when it isn't. Maybe the player did exactly as they should have, but the coach called a bad play. Maybe the kitchen mixed up the order. Regardless of where the error occurred, most of the blame ends up on last person involved.

The difference between credit and blame is in how we go about accepting them. For the most part we're open to receiving credit, and we tend to deflect blame. It isn't entirely our fault, we evolved to be that way. Credit means you helped the tribe, it provides status. Blame, on the other hand, means you did poorly. Of course, neither of those things are as true as they feel, but that's a part of the human condition isn't it? We place an extraordinary amount of value on the few variables we can control, and try not to focus as much on the myriad of those we can't.

Additionally, we're far more likely to give credit to things with faces and feelings. It feels good to give people (and animals) props. We're also more likely to assign blame on inanimate objects or the environment. (It doesn't hurt that they can't argue with us and take the blame in silence.)

What Does This Have to do With AI?

Great question, thank you for asking. As you're aware, AI doesn't have a face (yet) and doesn't have feelings. At least not in the way we understand them. Despite our desire to anthropomorphize AI, it's still far more "object" than "person" (though it feels like that's going to be a blurred line to cross when the time comes doesn't it?) and most of the wonderful work these inanimate objects are doing, are behind the scenes at the direction of, and in collaboration with, a number of remarkable people.

Which means that despite AI already doing some wonderful things; from discovering a new class of antibiotics to making easy to access tools to identify eye and skin disorders, cancers, and many other diagnostic feats. It is often the physician you'll thank for catching that disease before it became a problem, or curing that thing that previously had no cure. When an AI helps customize a curriculum, enabling your child to fulfil their potential, it is likely to the be teachers you give the majority of the credit.

Unfortunately, this also means AI will be an incredibly convenient inanimate scapegoat for anyone who needs to deflect blame. Certain flavors of CEO's already excel at this - the layoffs are the result of "market conditions", "the economy", "financial performance" and a myriad of other faceless factors, not their own mistakes or oversights. Artificial intelligence will be the latest of these factors - an easy to point to scapegoat to avoid taking responsibility themselves. This problem will only be multiplied by media, operating as it is, largely on an attention economy which incentivizes them to strike fear into their audiences so they watch, share, and engage more frequently with that "news."

The Takeaway

If you haven't already, you'll soon be hearing about all the terrible things AI can, will, and perhaps has, done. However, you're unlikely to hear nearly as much about all the positive things it does to help you. You'll attribute that to the faces in front of the machines - the doctors, teachers, mechanics, and thousands of others whose industries will be revolutionized by the (ideally ethical) implementation of artificial intelligence. The blame, however, will be handedly redirected towards AI. From there, it will be amplified by media outlets focused on converting fear into revenue, and finally that signal will be bounced around in echo chambers we until it has sufficiently distorted our view of the world into believing that the negative is the whole picture.

Don't get me wrong - the rollout of AI will lead to issues. It is our responsibility to find ways to implement these tools ethically to help insulate society from the culture shocks technological breakthroughs cause when not deployed with intention. Regrettably, our approach to deploying new tech is laissez faire, bordering on reckless. Which means it is up to each of us, individually, to remain aware of this distortion and resist blindly giving in to the binary "good" and "bad" thinking foisted upon AI.

###

1 - Last touch attribution bias is kind of a malaphor. It combines "last touch attribution," which is a marketing attribution model in which 100% of the credit for a conversion is given to the touch that occurred directly before the conversion (as someone with a background in marketing, I can assure this is not a great model to use and is incredibly short sighted) and "attribution bias" which is a psychology term for a cognitive bias that refers to the systematic errors made when people try to find reasons for their own and others' behaviors. Specifically, it focuses on the applications in which this results in perceptual distortions and inaccurate assessments of situations and events.

2 - re: "Lion's share" I understand the idiom originates with Aesop's Fables, but in the interest of scientific accuracy I think the idiom should be "Lions' share." They're pack hunters, so plural possessive is more accurate. And even outside of the fable, they eat the majority of their kill, leaving scraps for hyenas, dogs, vultures and other scavengers. This is just one of many molehills on which I am willing to die. 💀