Never mind the ridiculous videos of Mr. Rogers chatting with Tupac Shakur or Bigfoot vlogging, these AI videos seem real enough while fully intending us to know they are AI. What we are seeing now is an indistinguishable bending of real and fake with videos that are completely altering our ability to know what is real and what isn’t.
Voice mimicking was already almost perfect. I saw a video post today from a man whose dad called him to ask what their shared bank account password was. One problem: His dad died last year, he just hadn’t taken his name off of the account yet. He said it sounded so real that had his father been alive, he probably would have shared the password, thinking his dad forgot.
Now AI videos are just as good as AI audio and the combination of the two truly are steering us into a post truth era. People are sharing AI videos completely unaware that they are fake. Even news stations are getting it wrong.
Soon web sites will become bastions of truth. Want to know what someone actually said? Go to ‘their name’ .com or .org and see the actual video shared there. Anything else will be questionable. And wherever else the video is shared must be watched with skepticism. Subtle or overt, very important changes in a message will occur as a result of someone, ultimately anyone taking the original video and making an AI version that gives their message instead of the intended message.
Following specific domains, and maybe a handful of legitimate news channels, are the only suggestions I have. Legislation won’t keep up, and the fakes are just getting better. Essentially, find reliable sites and distrust everything else. Intuition and common sense won’t be enough.

Pingback: AI – Alternate Identities | Daily-Ink by David Truss