,
9 tweets,
2 min read
Read on Twitter
As soon as we say AI’s, as ‘agents’, have a moral responsibility, the risk is that people who actually have it (financers, politicians, designers, engineers, users) feel excused from it.
we should be really clear on what we mean by technology having moral agency. It is clear that technology has moral *influence*. Having a gun in hand makes you see things as shootable: this is basic affirdance theory /1
Having a bureaucratic system with in-built bias makes people behave biased, people follow the constraints of the system, we’re sheep in the IKEA./2
So there is agency in the sense of it being a factor - it plays it’s part and individual humans in the flow of everyday cannot simply escape it. It would be unfair to say to a human caught up in technological mediation O but you could have simply decided not to be mediated. /3
But this is different from saying that the gun or the bureaucratic IT system itself are moral ‘persons’ that can be held accountable. And the reason is that helding then accountable will not *change* anything. The change needs to come from designers and other decision makers. /4
So we can say ‘the system caused it’ but we cannot say: call the system and tell it to stop it and behave. What is more, if we grant the system status as moral agent in this ‘personal sense’ the people that actually can change things will no longer feel responsible/5
There is the real danger that talking about moral agency of machines has the effect we can hide behind it and forget to confront ourselves.
*holding them
@threadreaderapp unroll
