I think Google’s “Reimagine” feature (shipped with their Pixel 9 phones) marks the beginning of the democratization of AI-powered, Very Good Photographic Fakery. Soon, every phone vendor will ship with this or have apps that support this.
-
I think Google’s “Reimagine” feature (shipped with their Pixel 9 phones) marks the beginning of the democratization of AI-powered, Very Good Photographic Fakery. Soon, every phone vendor will ship with this or have apps that support this.
I think this is a watershed moment. Photographic fakery will be so good, and so cheap, it will be indistinguishable by most people from truthful photography. I’m sure we will adapt, and astute, tech-aware viewers will continue to be able to pick out fakes. But at this quality, it will be beyond the ability of the casual viewer. This is going to change things.
-
dataramareplied to Dave Rahardja (he/him) last edited by
@drahardja While some of the other examples are pretty scary, that bear, specifically, *looks* AI-generated. Look at the legs.
But yeah. Just under 100 years ago, Stalin needed to hire an airbrush artist who spent days of work to remove Trotsky from history. Ten years ago, a guy with Photoshop could do it in an afternoon. Now, a child can do it in ten seconds.
-
Dave Rahardja (he/him)replied to datarama last edited by
@datarama Yeah, like I said, it’ll still be possible to tell fakes from real, but it’s getting more and more difficult. And because this feature will be included free in a popular phone (other phones and apps can’t be far behind), we will have to deal with more potential fakes all around us. I think we are now at a point where the average viewer will not be able to tell fakes apart.
-
dataramareplied to Dave Rahardja (he/him) last edited by
@drahardja We've been there for a while already. At least if some of my relatives who still use Facebook are any indication.
This particular feature, it seems to me, is the *definition* of that Ian Malcolm quote from Jurassic Park. Sure, it's technically neat, but *WHY?!?!?!*.
-
@drahardja As one of my friends remarked: "Great - now we can't communicate with photos anymore."
-
Dave Rahardja (he/him)replied to datarama last edited by
@datarama I don’t think it’ll come to that. As with any fake-generation tech, the source of the image matters, and it matters much more today. Just as we trust reputable news organizations to tell us the truth, we trust our friends to send us pictures they actually took and have not doctored.
And we will develop a healthy sense of skepticism. We will continue to use fact-checkers and pause a bit longer before forwarding provocative images.
We’ll be fine, but more and more people who are unprepared will fall into a credulity hole. More and more will get manipulated and scammed before they learn their lessons (if ever).
-
dataramareplied to Dave Rahardja (he/him) last edited by
@drahardja He was being hyperbolic, obviously - but only a bit. Because you're right: I don't have any friends who would send me doctored photos *without* it being obviously and explicitly a joke. I'm more likely to trust a picture brought by a credible news organization than by some dude named Trump4GodEmperor88. And so on.
But the latter is the problem - photos have now become much harder to use for mass communication. We don't agree which organizations are credible, after all.