Your eyes are now officially unreliable.
AI imagery has crossed a line from “obviously synthetic” to “unapologetically believable”. This fundamentally shifts the issues from technical glitches to psychological blind spots.
The Future Is Appearing Faster Than Ever.
I first wrote about AI art back in 2019, when AI-generated self-portraits were suddenly everywhere.
By 2022, the technology had advanced enough to warrant a revisit.
This May, it leapfrogged again, when OpenAI’s GPT-4 went far beyond clever image generation into something more akin to actual art. It felt like a turning point, not just for what AI could make, but for what it meant.
Now, only six months later, we’re at yet another inflection point. “Seeing is believing” means a lot less than it used to. AI imagery has become so realistic that you often can’t tell whether what you are seeing was captured from reality or generated by AI.
That may sound like a small shift in aesthetics, but it marks a big shift in how easily AI-generated content can slip into our feeds and our beliefs unnoticed. Ultimately, this will have a profound impact on what we think, what we know, and how we make decisions.
And the quality isn’t limited to paintings or digital compositions — these generation capabilities apply to portraiture, video, and other convincing replicas of real life.
Do Your Eyes Deceive You?
Here is an example a content creator posted on X, showcasing two images created by Google’s Gemini AI. The one on the left was created by the original version, and the one on the right was created using Google’s new Nano Banana engine.
@immasiddx via X
The first has the classic look we associate with AI — shiny, airbrushed, and clearly AI-generated.
The latter looks like it came from a random Instagram post. The image’s texture and tone look real.
I didn’t choose this image because it was technologically impressive. Instead, it was how unremarkable the ‘better’ image looked — ordinary enough to blend into a feed without ever tripping your ‘this might be AI‘ alarm.
At the same time, Sora 2 is affecting video in the same way.
While you could use a tool to check whether an image or video is likely fake or real, what percentage of the population is doing that? And even if you did … we got to this level of quality fast enough that it’s hard to imagine what it will be able to produce in the near future.
Artificial Reality Has Real Consequences
Meanwhile, people are increasingly consuming and believing AI videos they find on social media or even mainstream programming (here’s an example from Fox News). Further, some people don’t care whether the image was captured live or generated (because they consider it all part of their narrative-building process).
This video of bunnies jumping on a trampoline fooled so many people that it inspired musician Oliver Richman to release a song about them.
via YouTube
This video might seem trivial because the it looks pretty unassuming and straightforward. But it shows that we are now on the other side of a gateway to artificial, augmented, or alternative realities … and the implications are enormous.
For example, automation bias and grandparent scams are becoming more common, capitalizing on these trends.
Real Eyes Don’t Always Realize Real Lies
AI is an incredible tool, but it can also reduce people’s willingness to question what they see. Some might argue that, even more dangerously, AI makes the least curious part of the population even less inquisitive.
AI-generated content is such a reliable crutch that many people now trust it at face value. For some, it has replaced friends and loved ones (not to mention search engines) as their go-to source for information or confirmation.
But you don’t have to be intellectually lazy to fall for AI anymore.
As AI imagery and video become indistinguishable from reality, the hard part is no longer generating convincing content; it’s preserving your ability to distinguish signal from noise.
Avoiding automation bias means staying skeptical of anything that looks too perfect or aligns too neatly with what you already believe.
Seeing Clearly In an AI World
Treat AI-generated content the way you would any unverified claim: look for the source, check for corroboration, and slow down before reacting. The easiest way to avoid being fooled is to build a habit of pausing and asking, “Who created this, and why?”
Another great strategy is to rely on multiple trusted signals rather than a single compelling image or clip.
In a world where fakes are frictionless, trust becomes more valuable, and critical thinking becomes a survival skill.
Leave a Reply