3 ways to prove you’re human online
As AI overwhelms the web, we will need a way to distinguish people from machines.
by Andrey Mir
In 2010, Eric Schmidt, then-CEO of Google, claimed that every two days, humanity was creating as much information as it had generated from the dawn of civilization through 2003 — 48 hours’ worth of texts, photos, articles, tweets, and other content added up to more than five exabytes of data, according to Schmidt.
Since then, generative AI has helped take our shift from information scarcity to information abundance to a whole new level — as of September 2025, we were generating more than 16 exabytes every hour.
AI-generated content now accounts for an increasing share of the information we generate — some estimates suggest that it could soon exceed human-generated content. As this trend continues, content made by humans could become relatively scarce, and because scarcity creates value, the human touch will become more valuable.
Long gone are the debates about whether AI can effectively mimic humans, though, and the better its content gets, the harder it will be to prove something was made by a human. Like Diogenes wandering the agora with a lit lantern in daylight, muttering, “I am looking for a man,” soon we will wander the uncanny valleys of the internet, searching for a human soul amid piles of AI garbage and hordes of AI assistants. How will we ever be sure we are reading something a real person wrote or speaking to an actual human being?
A new type of Turing test is needed: Instead of one that proves a machine is sufficiently “human-like,” we need something humans can use to prove their humanity to each other and the information ecosystem. We need proof of life, not proof of machine intelligence. So, how can you prove to others in the digital world that you are a human and the content you create is human-made?
Show human ID
Many online services and procedures already require users to submit a photo of their government ID, sometimes paired with a live shot of their face. This is currently the most reliable form of human verification we have for mass use. Still, people usually need a strong reason to go through this cumbersome process.
Some governments are pushing for citizens to get digital IDs — instead of submitting a photo of a physical ID, you’d use a government-issued digital one. Will these become the main method of human verification online? They may — and that wouldn’t be good news.
From healthcare and banking to working and socializing, more and more of life is happening online. If a government-issued digital ID is required for access to digital spaces, the government could not only track your online activity but also deny you access to essential services with a few keystrokes. The only thing worse than surveillance capitalism is surveillance socialism — the state controlling everyone’s existence digitally.
Digital IDs will, of course, be declared highly secure against identity theft. But if such theft occurs, the consequences could be catastrophic. As more of our lives depend on digital systems, a stolen digital ID means a stolen life.
Another way to verify identity is for services to ask users to confirm they are human through biometrics on their devices, such as a fingerprint or facial recognition. It is not a mandatory ID — just voluntary verification for human access when it matters. Whether this approach will lead to large, human-only platforms remains to be seen. Some additional incentives will likely be needed. But imagine a platform with proven human-only activity. In a couple of years, it could become a huge asset.
Show human effort
People tend to value something more if they see effort put into it. In cognitive psychology, this is known as the effort heuristic: people judge the value or quality of something by how much effort they believe went into it.
Industrial society taught us to value outcomes, not effort. A product had to work, be affordable, and signal our status — these were the things that mattered. Judging something by the amount of effort put into it was considered a fallacy. Still, according to studies, people value a poem, a painting, or a suit of armor much more when they are told that more time and effort went into making it. Effort spent becomes the central criterion for value when you don’t have or care about other criteria.
Now, the fallacy of judging by effort spent is reversing into a criterion of genuine value. As AI takes over marketing, a growing trend is pushing brands, producers, and creators to show not just the product itself — nothing special to see there anymore — but the human effort behind it.
“The plot of Mission: Impossible is almost as important as the subtext that Tom Cruise does his own stunts,” Rachel Karten writes in her essay “Evidence of effort: Why brands should make effort a character in their social story,” before delving into how to use the effort heuristic in digital marketing.
This theory offers a new marketing psychology. Whether you sell art, food, content, or any consumer goods, show the process, the team, and all the behind-the-scenes details — the elements that AI lacks. This idea is promoted as if we will soon live by two guiding mottos: “We humans work hard” and “We humans must support our own kind.”
Show human flaws
Aristotle said that art imitates nature (ars imitatur naturam), and for centuries, many artists sought to imitate their subjects as realistically as possible. Then photography was invented. With it, new media began taking over the arts’ function of representing reality. Artists began to care less about the accuracy of representation and pivoted to self-expression. Painters turned to playing with what the camera could not capture: perception, emotion, subjectivity, abstraction, inner vision. Realism gave way, in part, to Impressionism, Expressionism, Symbolism, Cubism, Abstractionism, Surrealism, and more.
Interest in imperfection, roughness, and asymmetry was already growing in the digital era before AI arrived. The trend was captured, humorously, by Left Shark’s awkward dancing during Katy Perry’s Super Bowl halftime show in 2015. Amid the incredible automatism of the stage production (and life in general), Left Shark’s choreography mistakes spawned a wave of internet memes that made Katy Perry unhappy — until her lawyers saw the value and attempted to trademark Left Shark’s design.
Perfection alienates; wrongness invites. The arts, and later marketing, learned to use this. Now it is becoming the test of human authenticity itself. The more perfect the machines become, the more valuable our imperfections will be. Evidence of life may reside precisely in the things machines once tried to eliminate: friction, asynchrony, irregularity, and imperfection. We used to prove our humanity by what we could create. Now we are pushed to prove it by what we cannot perfect. Authenticity will live in the margins.
Left Shark’s dancing may have accidentally revealed a prophecy. A dancer who forgot or failed to perform automatic moves became the most human part of the show machinery. In a world flooded with flawless machine output, the most human thing may be the slightly (or not so slightly) wrong thing: the awkward phrase, the imperfect gesture, or the visible underachievement.
Evidence of life
Every previous wave of automation — the printing press, photography, electric replay, etc. — eventually reinforced human expression, rather than diminishing it. In a sense, the automation of thinking by AI may follow the same pattern by forcing us to discover what is uniquely human and hold on to it. Alas, it is not intelligence or creativity — machines have long outperformed us in terms of intelligence, and creativity can be easily mimicked by modern AI. The criterion of self-consciousness, meanwhile, is of no use for the practical task of distinguishing between human and nonhuman digital activity.
So, proof of life in the AI era may come from three places: digital identification, a show of effort, and an embracing of flaws. Choosing digital IDs or biometrics could lead to surveillance socialism. Choosing effort could lead to performative “labor signaling.” And choosing to verify humanity through flaws could lead to a society that esteems errors, ugliness, and underachievement. Ultimately, as we decide how to prove our humanity in the AI age, we will be making not just a personal choice, but a civilizational one — and there are no obviously right answers.
 | A guest post by
|
What a beautifully uncomfortable piece.
Andrey names the paradox without fully sitting in it: we are being asked to prove our humanity to systems built by humans, for humans, now running humans. The Turing Test didn't fail — it inverted. We're the ones in the box now.
The three paths he offers are real, but each one carries a hidden cost he acknowledges honestly: IDs risk surveillance, effort risks performance, and flaws risk valorizing dysfunction. What he doesn't say — but implies — is that all three are forms of legibility to the machine, not expressions of selfhood to each other.
The deeper question isn't how do we prove we're human to algorithms? It's why have we agreed that algorithms get to decide?
Left Shark wasn't proof of humanity. Left Shark was a reminder that presence leaks through. You can't perform that kind of alive-ness. You can only be it — and then someone notices.
Maybe that's the fourth method he didn't write: stop performing humanity and just live it loudly enough that it bleeds into the digital record. The proof won't be in the ID, the effort, or the flaw. It'll be in the irreducibility. That's the thing that doesn't compress cleanly into a training set.
Andrey MIr....You misquoted in quotation marks, the words of Diogenes which wholly changes its meaning . The correct quote is "I am searching for an HONEST man"....I assumed it was likely a mistake by the printer and you'd appreciate this correction.
I enjoyed your work regardless...Thanks.