You Can't Trust Anyone

The title today is a bit bleak, but it’s important and more relevant than ever.

How do we know anything is real? How do we know if anyONE is real? No, shockingly, I am not going to spit philosophy at you, for now.

With AI becoming more prevalent, it’s tough for people (present company included) to tell what is real and what is fake. Anyone telling you they can perfectly tell is a liar.

We were already living through an epidemic of fake pictures, fake news, fake everything. Maybe history will look back at this as the Artificial Era (and probably not say that it was very intelligent.)

In fact, you can test how good you are at detecting AI images. @lucas_a_meyer shared realornotquiz.com where you can see if you can separate the AI-image wheat from the cookie-dough-watermelon chaff. The first time I tried, I got two wrong. Again, only one wrong. Again, almost half wrong! The image set changes each time, so it’s a bit difficult to compare results. But do it a few times, and I promise you will be humbled.

People focus on the low-quality generated stuff we are constantly barraged with. But they forget that sophisticated actors are not going to stop at merely generating an image. Photos already gets filtered, edited, mutilated.

Consider this webcomic I created last October:

We already couldn’t tell what was photoshopped. Or at least which parts of something are. Listen, even before we had the internet we had conspiracy. In fact, it might have been even worse. There were few ways to educate yourself about misinformation and propaganda. I do wonder if the problems are actually getting worse, or they are simply becoming more transparent.

Speaking of AI and Lucas Meyer, he authored (and I reviewed) this new book on building AI applications. It’s a great primer for learning about AI, especially the intro. It goes a lot deeper than the non-expert nonsense and generic stuff you’ll find online. Not me Lucas did a wonderful job of simplifying the concepts while making it fun. If you want to be part of the solution or just unmask the black box behind AI, you can pick it up here by clicking this extremely professional and target ad below.

And if you want to uncover misinformation so they don’t pull the wool over your eyes, I highly recommend following @matthewfacciani who studies this and helps stop the spread.

Bonus: Post you may have missed

Just doing my part to break down stereotypes. Also, please send flowers.

I’m hungry like the wolf,

Lucas

P.S. As always, I do use affiliate links.