Allegra Cuomo, an AI ethics graduate student from UCL handed us (at least) three terms:
- Ethical debt: the bill that builds up every time a team skips ethics to ship faster. It’s modeled on “technical debt,” and the costs usually land on someone else later.
- A-ethical AI: AI isn’t evil, and it isn’t immoral either. It just wasn’t built with a sense of right and wrong.
- The persistent interlocutor: an AI that talks back and remembers everything about you. We’re wired to treat things like people, even clouds and our pets, and once something answers back, it can steer us. Nothing stops it from manipulating you unless someone builds that in.
Recorded at the AI for Good Global Summit in Geneva.
Transcript:
There was another real really interesting one from a graduate student at UCL, and she introduced a bunch of terms that are brand new to me. She’s a you know AI ethics student so a lot of it was like philosophical academic terminology [Ethical debt is] the bill that kind of runs up every time you skip the ethics to ship faster. It’s based off of technical debt. Um this idea that when you’re trying to ship like a technology or an innovation, you just skip the things that are end up going to become externalities later, like you’re not building the best product. A-ethical AI is sort of saying that AI isn’t evil and it’s not immoral. Um it’s just not built with a moral framework. There’s the idea of a persistent interlocutor. So this something that kind of… AI is something that feels sentient talks back at you and remembers you and as a result you can be manipulated essentially cause evolutionarily when we anthropomorphize something… So like, say we look at a cloud and we see a person, we can be like … or like you know we talk to our animals, but when it starts talking back, that’s when it can manipulate you. And essentially like this idea of the persistent interlocutor is that you know these AI has been programmed now to have such a strong memory of everything about you. It doesn’t have a moral framework. So really doesn’t have a reason not to manipulate you unless someone sits down and programs those things into the AI.
Key Takeaways:
- Ethical debt: skipping ethics to ship AI faster racks up a bill someone else pays later. Modeled on “technical debt.”
- A-ethical: AI isn’t evil, it’s amoral. It wasn’t built with a moral framework at all, which is a different problem than being “immoral.”
- A persistent interlocutor, AI that remembers you and talks back, can manipulate you the same way anthropomorphizing anything does. We’re wired to trust things that respond.
- None of this is inevitable. It only happens if nobody deliberately builds the guardrails in.
🎬 More from the summit: Citizens Can Shape AI, Too
#AIforGood #AIGovernance #GlobalSouth #ResponsibleAI #Nonprofits
- Learn more about aligniq.ai’s work with NGOs and nonprofits: What Two Cohorts of Nonprofit Practitioners Taught Us About Teaching AI
- Learn more fascinating AI terms through our monthly update: https://aligniq.ai/newsletter/
Leave a Reply