Tim Requarth: Yeah, so memory is a little bit different than phone storage, where you have a photo and you can go look it up. And if you’re not looking at it, it remains the same. And when you pull it up, it’s the same every time. In memory, what’s interesting is it’s a replay, not a retrieval. And so, with each replay, it can be a little bit different. So, memory is inherently a little bit unreliable, which I think is something we would all recognize from our day-to-day experience.

Right. It’s really interesting. And I’ll be honest, I was surprised to hear this. … Let’s say you make a lot of these videos of yourself doing things like, you know, I had my friend who climbed Mount Rushmore, people make videos of themselves dancing in K-pop. People would make more mundane videos of them having a conversation with somebody about just a topic that they were interested in. So, there could be any number of ways. And imagine that you’re doing this hundreds of times. The memory of actually creating that video starts to fade. And the video itself starts to get recalled, but it gets kind of confused with whether you did it or not. And the more realistic or plausible it was, the more likely you’re going to mislabel it as something, not that I did with an AI app, but something that I did personally.

Yeah, and because it’s video, it’s something vivid when it comes up in your mind’s eye. And it says, “It’s vivid, so it’s probably more likely to be real.”

Yeah, I think that’s a great question. I mean, somebody can look at a photograph over and over again, and eventually it can insinuate itself into your memory. So this is not in any way about AI only. The difference is that AI can be very realistic and it can be infinitely generated and watched, and it’s pervading our lives in a way that is different than previous generations of media.

I find it concerning when a technology can trip up what I call root processes or very fundamental processes in the brain. And so, for example, we’ve already seen with social media that it encourages people to make comparisons of themselves to other people, influencers that maybe are projecting an unrealistic ideal of beauty, intelligence or success and it makes people feel bad. Now, imagine what you can do when you’re seeing images of yourself not quite living up to this reality. So, I think there’s a big potential for this kind of technology to drive or turbocharge some of the dynamics that we’re seeing in social media and take them to the next level.

You could or a company could, potentially.

I think, potentially, yes because I think that it can become internalized in a sense of identity of who we are, where when we really look at a real picture or our real actions in the real world, and they’re not living up to something that has been implanted, essentially, it’ll be hard to be yourself. It’ll be hard to break out of that.

Yeah, and that’s a great point. A lot of these things happen a little bit unconsciously and you have to reject them later in your brain and that doesn’t always happen, especially for younger and potentially more vulnerable users.

You’re right. That’s who we are. We are the collection of our memories, and we narrativize those to generate a sense of identity. And if that sense is contaminated, it’s gonna contaminate our sense of who we are and maybe what we can be. Is that gonna be neutral, positive or negative? I’m not sure, but the technology is capable of that. And you know, it remains to be seen what will happen with the next generation of self deep fake technology.