AI Image recognition: how good is it in 2025?

Some days I am jolted out of my own complacency.

Recently I was sent on one of those GenAI training workshops - I've been to many of them now as an IT consultant. Our clients and the whole business ecosystem seems busy getting up to speed on it. One the more interesting hands-on workshops had us test the image analysis capabilities of various models.

Funny how we take image recognition for granted now. I couldn't help be reminded of this xkcd cartoon from 2014.

They say humour doesn't age well, but this went semi-viral in 2014 – it particularly encapsulated the frustrations techies have explaining the art of the possible and the insane – nicely dissected in this contemporary Explain XKCD article. The alt text, where xkcd's author Randall Munroe likes to hide a hidden punchline, sums it up:

In the 60s, Marvin Minsky assigned a couple of undergrads to spend the summer programming a computer to use a camera to identify objects in a scene. He figured they'd have the problem solved by the end of the summer. Half a century later, we're still working on it.

Anyway, back to the corporate training day. Rather than feed it the image from our lab instructions, I grabbed a rather old and very random picture from my phone. This one was taken side of the road, no idea why, by a petrol (gas) station somewhere nondescript:

So I tried

The robot answered:

Eye-level view of a parking lot and surrounding area on a cloudy day.

The foreground of the image shows a wet parking lot, with traffic cones placed near what appears to be a designated area or entrance. The ground is dark from the rain. A white van is parked within the lot, facing slightly away from the viewer. A low stone wall and a metal railing are visible in the bottom left corner, seeming to separate different levels or areas of the parking lot.

Several signs are posted on poles of varying heights. These signs appear to relate to parking regulations, with yellow backgrounds and dark text. A blue post- mounted box, likely for mail or deliveries, is visible near the center.

The background features a line of trees and buildings. The buildings have chimneys and appear to be houses or small commercial structures. The trees are a mix of evergreen and deciduous, with some showing autumnal colors.

Beyond the trees and buildings, a range of hills or mountains can be seen in the distance. These are partially obscured by a layer of mist or low-hanging clouds, which contribute to the diffused light in the scene. The sky is predominantly overcast, with some lighter areas suggesting breaks in the cloud cover. There’s a subtle hint of light, possibly from a setting or rising sun, creating a hazy, luminous effect behind the distant hills.

Preliminary grade: Holy shit!

Let's see how far we can push it. There are some levers we can play with.

Try this:

• Signage: The yellow parking signs with specific time limits (e.g., "90 minutes") and the general design are common in the UK and Ireland.
• Van: The white van appears to be a Ford Transit, a very popular model in the UK. The number plate format, while blurry, seems consistent with a UK or Irish registration.
• Architecture: The chimneys on the buildings and the style of the stonework are suggestive of British or Irish construction, possibly from the early to mid-20th century.
• Landscape: The rolling hills in the background, along with the mix of deciduous and evergreen trees, suggest a temperate climate similar to that of the British Isles. The general cloud cover and rainy feel of the image are also consistent. It's difficult to be more specific without seeing sharper details of the license plate or any business signage. But these clues strongly point toward somewhere in Great Britain or Ireland.

Let's jack up the

Several clues suggest this photo was taken in the United Kingdom or Ireland:

• Signage: The yellow and blue parking signs with specific time limits (e.g., "90 minutes") are a common style in the UK. The font and overall design are typical.
• Van: The white van appears to be a Ford Transit, a very popular commercial vehicle in the UK. While available elsewhere, its prevalence makes it a hint.
• Architecture: The chimneys and slate roofs of the buildings in the background are suggestive of UK architecture, particularly older styles.
• Weather: The overcast, drizzly conditions are typical of the UK and Ireland.
• Landscape: The rolling hills in the background, while not definitive, are consistent with the terrain found in many parts of the UK and Ireland.

While not conclusive, the combination of these factors strongly points to the UK, possibly Scotland or Ireland given the hilly terrain. Ireland is less likely due to the car's license plate, which appears to be a UK format.

Definite points deducted for the very clearly mis-identifying the rear car plates as British (which are yellow), rather than Irish (which are white). A mistake you wouldn't catch if you didn't know it.

I don't know what I find more unnerving: how much it got right, or in how confidently it bluffed its way through a very obvious (if you know) error.

All of which reinforces my view that these are really powerful tools that may accelerate our work. But you need to check its work carefully before relying on anything it produces.

TLDR: Holy shit! Also, B-minus.