2 ms·
I agree with your main point that these are not new risks. Deepfakes have been available since Photoshop existed. I can create an photo (or even video) of you
by takinola 3y ago
I agree with your main point that these are not new risks. Deepfakes have been available since Photoshop existed. I can create an photo (or even video) of your politician of choice doing something untoward without using AI. What is the additional risk that AI brings to this scenario?
- ggm 3y agoPutting the high Q product into the hands of anyone, and established beliefs. The general trend is "but the computer said so, it must be right" as in google maps misdirection into swamps. How can anyone wind up in that place? Specifically photoshopping, no: thats been around for ages. Movies? Less such. They're pretty terrible without AI assist for smooth blending, and voice faking. Mainly I'd say upscaling from Photoshop to believable movies, and sheer volume of materials are the probable risks. But its incremental not revolutionary. And this is only specifically this one niche, right? Overall in the scale of things, I'd say the risk is misplaced belief overall. Applied to different contexts.