Interaction Designers have always shaped how people experience technology, but with AI, I feel like that responsibility deepens. I think we can all feel it now, especially since this is such a large and growing topic in our field. I think a lot of designers, including myself, realize we aren’t just designing interfaces, but also designing relationships between humans and systems that can actually infer, predict, and act. I feel as though there are so many ethical responsibilities for us as designers now and especially with AI and its integration with our work.
For starters, I think one huge thing we must understand as designers is that we cannot forget to respect user agency. We simply cannot start designing away choice. Also, in an AI context, though systems can suggest recommendations, it is important that we don’t assume anything about the user. I think similarly with AI, there’s this temptation to optimize for “frictionless” experiences by filling in the blanks, but again, I think we should stray from adopting this or letting this continue, as this also removes agency. On that note, as designers, we should definitely enable users to undo or opt out of decisions. It’s a relatively small decision choice but it adds so much more power back to the user.
I think with our own designs and especially with AI, we should help users understand what exactly the system or AI is doing and why, without overwhelming them. I feel like I always see two extremes these days of providing way too much transparency to the point of cognitive overload, or not seeing enough transparency. I think we, as designers, should use plain language always and be transparent when results aren’t guaranteed for whatever reason. One last huge thing that I am seeing more and more, and I think we should nip this in the bud, is that we should definitely limit or even fully stop anthropomorphism in our systems. Doing this implies intelligence, intention, and care when, normally, that literally isn’t possible from such a system.
Lastly, I think especially with AI, on the note of transparency, we as designers should encourage the users to anticipate misuse, bias, exclusion, and any other unintended consequences. I think with some systems, and especially with AI, there can be a lot of faulty information being shared, and we don’t want to perpetuate any misinformation. I think that ethics aren’t just about setting good intentions, but about anticipating or foreseeing any possible damage that could arise.
These are some ethical responsibilities that I think we designers should uphold, especially in the age of AI.