What Is That? How AI Sees the World
One of the things that fascinates me about artificial intelligence is how it doesn’t just write or crunch data anymore — it can actually look at images and try to tell us what’s in them.
Think about that for a second. For you and me, recognizing a dog, a tree, or a coffee mug is second nature. We don’t even notice how complex that process really is. But for a computer? That’s insanely hard. And yet, here we are in 2025, and AI can do it — sometimes frighteningly well, and sometimes in ways that make you laugh out loud.
That’s exactly what I put to the test in Episode 13 of AI for Everyone. I tossed a bunch of photos at AI and asked the simplest question: “What is that?”
How AI “Looks” at Things
The funny part is, AI doesn’t actually see in the way we do. There’s no mental picture forming in its “mind.” Instead, it’s analyzing patterns: shapes, edges, colors, textures, and then cross-referencing them against millions of other images it has been trained on.
That’s how it can glance at a golden retriever and confidently tell you, “That’s a dog.” Pretty wild when you stop to think about it.
Of course, the cracks show up when things get less obvious.
When AI Gets It Wrong
I had some great examples where AI nailed it. Everyday stuff like a chair, food, or plants — no problem. But then I threw in a photo of a sculpture from a local park. Instead of calling it “art” or “a statue,” it insisted it was… “playground equipment.”
Close, but no cigar.
And honestly? That’s part of the fun. The mistakes are hilarious, but they also tell us something important: AI is smart, but it doesn’t understand. It’s guessing. And sometimes, those guesses are way off.
Why It Matters Beyond the Laughs
Now, you might be wondering — who cares if AI can tell the difference between a statue and a jungle gym? Well, this same technology is showing up all over our lives in ways that are actually useful:
- Travel: Point your phone at a landmark and AI can tell you what you’re looking at.
- Shopping: Snap a photo of something you like and instantly find it online.
- Safety: AI is scanning bags at airports or reading X-rays at hospitals to spot things humans might miss.
- Accessibility: For people with vision impairments, AI can describe what’s right in front of them.
So yeah — me joking around with AI and saying “What is that?” is entertaining, but the underlying tech is pretty serious.
Humans + AI: Better Together
Here’s the bottom line: AI is fast, but it’s not flawless. A weird angle, tricky lighting, or abstract art can throw it completely off. Humans bring something AI can’t — context, common sense, lived experience.
That’s why I don’t see this as a competition. It’s more of a partnership. AI can do the heavy lifting at scale, and we can provide the judgment. Together, that’s powerful.
What I Learned
Honestly, this little experiment reinforced something I’ve believed for a while: the best way to learn about AI is to actually play with it. Don’t just read the headlines. Mess around. Test it. Laugh at its mistakes. See where it shines.
That’s the spirit of AI for Everyone. I want to make this stuff feel approachable, not intimidating.
Check Out the Episode
If you want to hear the full story — and get a few laughs at AI’s stranger guesses — check out Episode 13 of AI for Everyone:
🎧 Watch or listen here:
And if you’re curious to try AI for yourself, grab my free Beginner’s AI Starter Kit: aishow.io/free.
Final Thought
At the end of the day, asking AI “What is that?” isn’t just a silly game. It’s a window into how machines are starting to “see” the world. Sometimes they nail it. Sometimes they fail spectacularly. But every experiment teaches us something.
And if nothing else, it’s a good reminder: the future is coming fast, but it doesn’t have to be scary. Sometimes it’s just plain fun.
Did you know? Joel delivers AI keynotes for organizations that want their teams to actually understand and use AI. Learn more.
