CAN ALGORITHMS TRULY UNDERSTAND US?

Imagine you’re sad, and your phone decides, “Hey, you know what would cheer you up? An ad for a funeral home!” This is the magic of AI trying (and completely failing) to understand human emotions. Algorithms are everywhere: they recommend movies you may like, they guess our moods, or they even suggest we reconnect with an old friend.

But how well do they understand us?

When you Google “how to fix a sink,” suddenly, everyone that knows you is convinced you’re a professional plumber. When you buy a romantic gift, every website starts assuming you’re getting married. These are strange things! But if you pay attention, sometimes they can happen. 🧐

Can AI read our emotions? Well, it makes assumptions, but its predictions are biased in many cases: a “fake smile” at a boring work meeting can be classified as a happy face, and reading a menu without glasses can spur a model to “see” a sad face. 🥲

Have you ever talked about something once and then seen ads for it everywhere?

If you’re looking to buy a kayak, BOOM: ads for kayaks on every website for the next months (even when you switched to ping-pong). If you’re hungry and craving sushi, a couple of seconds later, “Best sushi places near you” will pop up somewhere.

AI is improving but still struggles with the complexities of the (unpredictable) human behavior. Models are amazing at finding recurring patterns, but they don’t truly understand us: kind of like a toddler trying to explain taxes. 😅

So, next time your phone suggests a breakup playlist before you even break up, simply remember: the robots are learning, but they still have a long, long way to go.

Please share this with friends who continuously say, “AI will take over the world.” At least they can prepare. 😉

That’s a wrap! Now go impress your friends with your AI wisdom.

Peace

The Neural Chaos