The Consumable Human

The Agora Meets the Algorithm

Long before there were MySpace, Facebook, Reddit, and many more, there was the agora. In ancient Greece, the agora was the public gathering place where people bought and sold goods, discussed politics, exchanged ideas, argued, listened and, inevitably, tried to persuade one another. The notion of people gathering is not something social media hatched.

Socrates famously tangled with the Sophists, teachers associated with rhetoric and the ability to make an argument effectively. Their inclusion here is not an accusation. Persuasion is neither inherently good nor bad. A teacher explaining a difficult lesson, a coach calling a play, a candidate making a case for office, and a salesperson describing a product are all hoping, in different ways, to produce some result in another person.

We have simply become much better at it. And by better, I mean creativity joined each technological breakthrough. The public speech eventually gained newspapers, billboards, radio jingles, television commercials and increasingly sophisticated advertising. Each new medium expanded the audience, while surveys, focus groups, circulation figures and statistical tools such as Nielsen ratings helped measure how that audience responded.

But those methods largely depended on samples, estimates and aggregate behavior. Even excellent statistical analysis could not observe every individual in the audience.

Then the agora went digital.

Curiously, the word algorithm traces back to the Latinized name of the Persian mathematician al-Khwarizmi. Today it broadly describes a defined procedure or set of rules used to produce a result. That result is something to get back to in a bit.

Put those two words together and something historically unusual appears. The agora gave us a place to gather, communicate and persuade. The algorithm gave the digital agora something its predecessors never possessed at this scale:

The ability to measure the people standing inside it.

The Scoreboard Nobody Sees

The old agora generally required two players: someone communicating and someone listening. The digital agora introduced a third—the platform standing between them.

Each time we open an app, we press Start.

Consider the resulting exchange as a three-player video game. The creator posts something and receives a visible score: views, likes, comments, shares, followers and, eventually for some, revenue. Those numbers influence the creator’s next move. A popular post encourages another like it. A dud quietly suggests trying something else.

The audience plays too, although its controls look different. Scroll. Stop. Open. Watch. Replay. Like. Comment. Share. Follow. Click somewhere else. Leave. Come back tomorrow.

The platform can observe those moves at remarkable detail. It can potentially know what appeared on a screen, whether someone stopped, how long they stayed, how much of a video they watched and what they did afterward. Different signals can be weighted differently within recommendation systems, creating something like an invisible scoreboard far more complicated than any game we would actually want to play.

The third player is therefore watching the other two.

Creators adjust what they produce according to the engagement they can see. The platform adjusts what it presents according to engagement it can measure. The audience responds to what appears, creating another round of information for both.

Nobody needs to be forced into the game. We press Start ourselves. Unlike the ancient agora, our actions no longer disappear into the crowd.

Engagement isn't something mysterious happening inside the computer. Engagement is human behavior translated into numbers.

So I decided to play with the numbers. I mean, are you surprised? You’re not surprised. I can’t imagine you’re surprised. Ok, I’ll stop.

So I Messed With It

Once I began wondering about the scoreboard, curiosity got the better of me. My Facebook posts had generally been limited to friends, so I deliberately began posting publicly. I wasn't interested in hiding something behind privacy settings and then marveling when the internet appeared to know about it. If I was going to play with the system, I wanted to give it something obvious to see.

Last fall, I went on what can reasonably be described as an extreme science kick. Physics, cosmology, technology and assorted trips down scientific rabbit holes filled my posts. Before long, my recommendations seemed to follow. Einstein quotes appeared alongside fascinating scientific graphics—some insightful, some perhaps a little more confident than accurate—and plenty of science reels.

Then I changed subjects.

I began posting heavily about faith, grounded naturally in my own understanding through Christianity. The recommendations shifted toward religious reels and pages. Later came adventure, and suddenly movie scenes and variations on the Hero's Journey seemed plentiful. By then, I actually found myself missing the science content, particularly StarTalk.

Eventually my posting wandered into romance, love and true love. Thank you forever & ever Mr. Rob Reiner for giving us all true love on the silver screen. I'm forever a hopeful romantic, so I have no complaints about the rather uplifting corner of the internet that followed.

None of this was a controlled experiment. I continued watching recommendations as I normally would, and I cannot know which signals Facebook used or how they were weighted. I wasn't trying to publish a scientific paper.

I wanted to see whether changing what I publicly expressed appeared to change what the digital environment presented back to me.

It did.

Which left a better question:

If our expression helps shape what is subsequently presented to us, what kind of feedback loop have we created?

The Mirror Has a Business Model

The algorithm doesn't need an opinion about you. It only needs a measurement of what keeps you there.

Because your attention has economic value. Something I’m sure we’re all aware of these days.

In 2025, Meta reported nearly $201 billion in total revenue. More than $196 billion of it came from advertising. Advertising is not merely something that occasionally interrupts the experience; economically, it overwhelmingly funds the experience.

That makes engagement particularly valuable. Another reel watched, another post opened or another few minutes spent scrolling creates additional opportunities to present advertising while generating more information about what holds attention. There is even growing research examining how problematic social-media use interacts with the brain's reward systems, although reducing that science to a simple claim that social media is “just like a drug” would overstate what we know.

There is also an interesting imbalance in our three-player game. A creator might supply the photograph, joke, essay, rant, expertise or ridiculous video that keeps someone watching. Successful creators can certainly make money, but Meta's own monetization terms make clear that participation and payment depend on eligibility, policies and access to its monetization tools.

Millions of ordinary posts can still attract attention without their creators receiving a share of the advertising revenue surrounding the larger experience.

Meanwhile, something else is happening.

The creator's work is consumed. The audience's attention is consumed. The resulting behavior—watch, skip, like, share, return—provides information that can help determine what gets presented next.

The mirror isn't merely reflecting us.

The mirror has a business model.

Who Exactly Are You Following?

A few months ago, as my public posts began receiving more views, Facebook asked whether I wanted to become a Digital Creator. I was curious about the monetization possibilities—who wouldn't be?—so I said sure.

For the record, my earnings currently total exactly zero dollars. Hey, when it starts pouring in, I promise more paying it forward will ensue. That’s a promise.

The designation did make me more curious about the content filling my feed. Consider something ordinary: a clever saying on a colored background, relationship advice, an inspirational paragraph or a beautiful photograph accompanied by a few thoughtful sentences. It might receive thousands of reactions, but who actually made it?

Perhaps someone wrote every word. Perhaps AI helped polish an original idea. Perhaps the creator and an AI went back and forth twenty times before either was satisfied. Perhaps someone requested fifty variations and selected the best one. Or perhaps the words, image and accompanying posts were generated with almost no meaningful human involvement.

Increasingly, the finished post alone may not tell us.

That creates a spectrum from human-created to AI-assisted, human-directed and, eventually, largely synthetic content. More importantly, generative systems can produce such material at enormous scale. An individual or organization no longer needs to create one post when technology can help produce dozens, hundreds or more, potentially far removed from whoever ultimately wanted the message distributed.

Recommendation systems can measure what performs. Generative systems can create variations of what performs. Those capabilities need not even belong to the same organization for the combination to matter.

Humans also tend to look toward other people when deciding what deserves attention, seems credible or appears widely accepted. Close your eyes, wait, open them and read this. You’re with your family during the holidays and Aunt Charlotte fills you in on all the celebrity gossip. Ok, now you can close your eyes a bit and think about that. Hopefully, you have an Aunt Charlotte. In psychology, this is often described as social proof. Online, we encounter versions of it constantly through likes, reactions, comments, shares and follower counts.

If synthetic accounts can also comment, react and appear to participate, something more fundamental changes. Artificial content is no longer the only thing being generated. The apparent human response surrounding it can be generated too.

That creates the possibility of synthetic social proof: not simply creating the message, but creating the appearance of a crowd responding to it.

Which leaves two uncomfortable questions:

Is the crowd real? What determined which crowd we would see? What am I doing here? When is Gamora?

The Consumable Human

None of this eliminates human agency. We can close the app, ignore the advertisement, scroll past the reel, disagree with the crowd or delete the account entirely. Persuasion changes probabilities; it does not erase choice. The behavior remains ours.

But when we choose to remain, another behavior can be measured.

That distinction matters because the digital agora requires both a continuing supply of content and a continuing supply of attention. Successful creators are given tools, audiences and eventually financial incentives that can turn posting into an all-day profession. Generative AI potentially accelerates that supply dramatically. Imagine, as one simple example, thousands of fans each believing they are having a personal exchange with a celebrity because an extraordinarily capable language model can maintain those conversations simultaneously.

The scale itself creates another concern. Socrates questioned persuasion that could succeed without necessarily being anchored to knowledge or truth. When content can be produced almost instantaneously, the distance between persuasion and truth need not be intentional to grow. Verification takes time. Generation increasingly does not. And in a social environment, perception can begin influencing behavior long before anyone establishes reality.

Meanwhile, the cycle continues. The person consumes the content. The platform measures the consumption. That measurement can help determine what comes next. The person consumes again.

For years, we have heard some variation of the idea that if a digital service is free, we are the product. Perhaps “product” no longer adequately describes the relationship. What these systems compete for is our attention itself: another click, another view, another comment, another minute. Unlike many consumable resources, human attention even replenishes. Tomorrow offers another supply.

Creators supply more reasons to return. Audiences return. Behavior provides new measurements. The cycle begins again.

The agora allowed people to persuade one another. Advertising scaled persuasion. Digital platforms made response measurable. Algorithms made that measurement adaptive. Generative AI potentially makes the material responsive too.

Socrates never had to ask what would happen when the public square could learn which argument, image, joke, face or story was most likely to keep each person standing there.

We do.

The algorithm doesn't need an opinion about us. It only needs to know what keeps us in the agora.

Maybe it’s worth asking:

When we look around the digital agora and see what appears to be the crowd, how much of what keeps us there is something we chose—and how much was chosen because it learned we would stay?