AI Explained by a Comedian

Ever feel like your phone might secretly be smarter than you? Comedian Lieven Scheire thinks so too, and he spent an evening making us laugh and learn about it.

Lieven started off with a nod to the past, bringing up Clippy, that old Microsoft paperclip assistant that was more annoying than helpful. It’s a funny reminder of how far we’ve come. Now, we have AI like Siri and Alexa who, despite their hiccups, manage our calendars and play our favorite songs without us having to lift a finger.

He then broke down something called neural networks, which is the brainy stuff behind AI, using dice as a prop. Imagine throwing dice to decide your every move; that’s sort of what these networks do when learning from data. It’s a bit of trial and error and a lot of learning from mistakes.

Speaking of smarts, Lieven pointed out that our smartphones have way more computing power than the computers used for the Apollo 11 moon landing. We use this incredible power mostly to stream videos and play games, but it’s also behind the smarter applications like real-time language translation. That means when you’re traveling, you can point your phone at a menu and get it translated on the spot—a big leap from our old phrase books.

Lieven didn’t shy away from the tricky bits of AI, either. He explained the “black box” of AI, which is tech-speak for not always knowing how AI comes up with its conclusions. It’s a bit like baking a mystery pie—sometimes you don’t really know what’s going to come out of the oven.

And yes, AI does mess up, especially when it’s trained on flawed data. For example, an AI might start thinking all snowy landscapes have to include a husky or a wolf because that’s what it sees most often in pictures. It’s funny until it leads to bigger mistakes, and fixing this is part of the challenge.

Ending on a curious note, Lieven talked about how those irritating Captcha tests on websites help improve AI. Each time we identify buses and crosswalks, we’re training AI to recognize these objects better.

Through his talk, Lieven showed us that AI might seem complex and a bit intimidating, but with a few jokes and stories, it’s a lot more accessible. And as much as it’s capable of extraordinary feats, it’s still not above making a silly mistake or two—kind of like humans, right?

Thank you, Lieven, for making AI a little less scary and a lot more fun!