Season 1. Episode 21. Unscripted with Nico Ianeselli

So now it should be obvious to most people, but you're from Italy, right?

Yes.

so uh as a country you know because we're here in the us like we see what our attitudes are of pro and for ai uh how would you describe your country's attitude towards ai in general

so there is a lot of controversy you know that america advanced europe regulates right so the path to the future uh i think is bright for both sides but i think europe is advancing at that way slower pace because they want to make sure that everything is human focused which is absolutely correct you know but they sometimes tend to over regulate things and instead of giving empowering for example engineers like in the us they try to trap engineers and then they box them against the wall and you know and then they're like oh no you know we need to change it we need to move so suddenly they do a You turn and then at the end of the day, they lose a lot of time.

That's why I prefer what we're doing here, you know?

So, I mean, how do you think they can balance that right regulation and control with innovation?

I think what we're doing at The Society Man is the right thing to do and we're exactly in the right time to do it, you know?

Well, we debated a lot, right? You know, we talk a lot about the ethics and what should happen.

Still unclear though, right? Both here in the States as well as internationally.

I mean, yes, you see certain countries in the EU moving forward with some sets of policy frameworks, but Compared to normal technology or governance, I mean, AI is moving much faster than any policy position, right, can keep up with.

So what do you think the right framework is? I mean, how do you keep a balance that applies ethical guidelines but not necessarily constraints? At the same time, how do you support innovation?

Absolutely.

I do believe that we need to always have human in the loop, you know, in every process that we have, because we need to maintain control of everything that is happening.

Sometimes we see rogue AI or shadow AI, and then we need to backtrack what it's doing, why it's doing that, and if it was a mistake in the programming or if it was like a deliberate choice, which is going to be very, very interesting from a futuristic perspective, but at the same time can be extremely scary because then it's going to have unregulated choices, you know, and sometimes it's difficult for us humans to understand why it went away.

So I think that creating some boundaries and then, for example, you know, having the possibility to do forensic preservation for AI when you can put the AI on a stasis, on a cell, you know, and then investigate what happened in order to avoid this to happen in the future.

Obviously, thinking of the worst case scenario, right? 90% is good, 10% is bad, but that 10% can be very bad.

So we need to constrain those leashes.