Thanks for sharing this Andy! This isn’t something I had really thought about before but I think it’ll be super important going forward as this type of deception becomes more common.
Yeah, and this just one ... what I'll call adversarial use case that can come out of this tech. There are more out there that can get dangerous really fast. With this, plus other deep-fake technology, I think we'll a whole new market pop-up of tech to authenticate person-to-person interactions. Imagine an authentication software where at the beginning of a zoom meeting, everyone enters their authentication key just to rove that they are who they appear to be. Or a having your CEO give you an authentication key when they make a request for you just so you can be sure they are actually the CEO.
.... This might actually become a whole column as I think through it :-) I had another conversation at AIGG about how we'll need to start "signing" works created by human beings, to be able to differentiate content. from humans versus algorithmic content.
It’s so interesting. It’ll be neat to see how AI-generated content detectors develop and if they can keep up with AI content being generated. My guess is probably not.
Thanks for sharing this Andy! This isn’t something I had really thought about before but I think it’ll be super important going forward as this type of deception becomes more common.
Yeah, and this just one ... what I'll call adversarial use case that can come out of this tech. There are more out there that can get dangerous really fast. With this, plus other deep-fake technology, I think we'll a whole new market pop-up of tech to authenticate person-to-person interactions. Imagine an authentication software where at the beginning of a zoom meeting, everyone enters their authentication key just to rove that they are who they appear to be. Or a having your CEO give you an authentication key when they make a request for you just so you can be sure they are actually the CEO.
.... This might actually become a whole column as I think through it :-) I had another conversation at AIGG about how we'll need to start "signing" works created by human beings, to be able to differentiate content. from humans versus algorithmic content.
It’s so interesting. It’ll be neat to see how AI-generated content detectors develop and if they can keep up with AI content being generated. My guess is probably not.