The algorithm gives us what we want, not what we WANT to want
By Jennifer Dziura
16 min read
If you want a conversation starter that isn’t “So, what do you do?”, try the startling:
“What does your YouTube algorithm like to show you?”
This question is certainly too personal for some gatherings, but you will gather much more varied answers than you would with “Seen any good movies lately?”
Movies, you see, are things we watch publicly. Even with CGI and the specter of totally artificial actors, movies still cost millions of dollars to make, and so rarely target our most niche interests.
YouTube videos, though, are tiny movies we watch privately, and thus questions about their subject matter may be perceived as invasive.
YouTube watching is, more often than not, a vice.
Among those I have asked “So, what does your YouTube algorithm think you want to watch?”, I have received an array of answers I never could have imagined.
Someone whose movie watching tastes are pretty mainstream replies “Oh, mostly those pimple popping shows.”
You discover that your father, who does not fish, watches videos of other people fishing.
The nicest lady you know is fed an endless stream of cruel yet extremely well-funded pranks.
Some of us are watching police dashcam footage of high speed chases, actors re-enacting medical cases strange enough to make it into medical journals, a pet groomer who fat-shames large cats while bathing them, or – among young people – a wordless, endless animated series called Skibidi Toilet (“The series follows a war between toilets with human heads coming out of their bowls and humanoid characters with electronic devices for heads”).
Occasionally, you get someone who says “Oh, BBC history documentaries!” and you wonder if this is the truest answer; do they have “Art of Eternity – The Glory of Byzantium” open in one tab and pornography open in eight more?
The default assumption in the YouTube question is that the answer will be at least a little shameful.
This is because the YouTube algorithm – like many other algorithms that shape our lives – has no respect for virtue.
It does not, like Socrates, distinguish between higher and lower pleasures1.
It shows you what you want – not what you wish you wanted.
It entices your worst aspects, then feeds them, like a blood supply to a tumor.
Outside of Philosophy, the Word “Virtue” Is Mostly Used to Talk About Women’s Sex Lives
This is a shame.
For Aristotle, virtue was the idea that – while following ethical rules is better than not doing so – the key to human flourishing is cultivating in yourself the character traits that make you want to behave ethically.
A virtue, according to Aristotle, is a “stable disposition” that makes you a good person and enables you to live well.2
If you’ve heard the word “virtue” in popular discourse, it has probably been about women’s chastity, which of course I will assert is, by itself, neither good nor bad. I think there’s a place for virtue in sexuality, though. If virtue is genuinely wanting things that are good for you and for others, it could for some particular person be virtuous to take hormone replacement therapy to increase their libido, or it could be virtuous for another person to swear off pornography if it is compromising their marriage. But it would probably also be possible to concoct an example in which more porn would be virtuous3, such as if you could use it to shape your own desires such that you now want things that are good for you and for others (not the most common use case, clearly).
That is, determining what is virtuous is not as simple as imposing a set of religiously-motivated strictures on women and enforcing them with shame.
The classical idea of virtue describes a practice you get better at with time – act courageously until you become naturally courageous – not a set of rules that permanently marks you if you violate them.
So, I mention the idea of virtue here because I think it perfectly describes what the algorithm erodes, wearing us down over time.
Would You Like to See Another Teen Killer Sentenced to Life in Prison?
I can’t write an article like this without revealing my own YouTube vice – a vice I doubt I ever would have even discovered without the algorithm randomly sprinkling it into my suggestions, noting my prurient interest, and sending more, stat.
“5 Guilty TEENAGE Convicts Reacting To Life Sentences” is a tragedy on a variety of levels – that teens are committing enough murders to create an endless stream of content, that minors receive life sentences in America, that these reactions are made public, and that two thousand years after the death of Aristotle we have created a machine to entice people to watch them.
This is the beginning of a rabbit hole including “Killer Husband Realizes Wife Is Still Breathing,” “Child Pred Gets Caught In Front of His Parents,” “Police arrest fleeing child rapist minutes before flight takes off,” and “Man’s execution-style murder caught on camera in North Miami Beach #Florida” (did we really need the hashtag?)
I think the first half dozen videos I watched gave me the general idea: most criminals are pretty stupid. The cops interrogating them don’t need TV-worthy twists or cutting-edge forensics. I’ve seen a few examples of incredible patience on the part of an officer making a criminal feel at ease before easing him into admitting that his wife’s body is in the deep freezer. I was impressed, but it was not a good use of my time. The little bit of knowledge I gained from this did not make me a better person. I did not “flourish.”
There Is More Than One Kind of Wanting
So, do I want to watch more videos like this? I do not. Will I click on them when they appear in the algorithm? Sometimes.
I say I do not want the videos, and yet YouTube’s data shows I clearly do want the videos. Am I just lying?
And here comes the concept of virtue, the Greek kind.
It is not a lie, saying I don’t want the thing when I do, in fact, want the thing. If you ask a recovering alcoholic if they “want” a drink, the answer may be: Yes, always. Every second of the goddamn day. But also: No, why would you even offer this, you know I am a recovering alcoholic, please do not ever offer me a drink again.
Both things are true. There is more than one kind of wanting.
If you want to watch something inimical to human flourishing, but you don’t want to want to watch it, you may take certain measures. You can go, as they say, “touch grass.” You can employ willpower, which although out of fashion, still works sometimes.
And on YouTube, you can mark individual videos “Not interested,” or you can up the ante by selecting “Don’t recommend channel,” with the intention of blocking a whole array of similar videos.
However, the algorithm weighs actions over declarations. Which makes a lot of sense if the algorithm is created by a corporation trying to maximize watch time in order to show you more advertising.
Research from Mozilla in 20224 discovered that “YouTube’s user controls have a negligible impact on preventing unwanted recommendations, leaving people at the mercy of YouTube’s recommender system. As a result, YouTube continues to recommend videos that people have clearly signaled they do not want to see, including war footage and gruesome horror clips.”5
That is, YouTube simply ignores us when we ask not to see certain content:
The most “effective” user control was “Don’t recommend channel,” but compared to users who do not make use of YouTube’s user controls, only 43% of unwanted recommendations are prevented — and recommendations from the unwanted channel sometimes persist. Other controls were even less effective: The “Not Interested” tool prevented only 11% of unwanted recommendations.
A brief comparison, if I may:
In my forties, I no longer metabolize alcohol as well as I did when I was younger, and as such, I have a “one drink maximum” in bars and restaurants.
If you also wish to limit your public alcohol consumption, try this: When you order a drink at the bar, add “…and I can close this out anytime, I have a one drink max.”
The bartender has no way to know if you are casually trying to improve your sleep quality, or if a second drink will turn into a fifth drink and thereby transform you into a screaming nightmare who will have to be ushered out. They don’t want the second one.
As such, this tactic has had a 100% success rate. Every single bartender I have ever said “one drink max” to has refrained from even offering me a second drink.
In a virtue contest between bartenders and YouTube, bartenders win hands-down.
Amusing Ourselves to Death
In the 1985 book Amusing Ourselves to Death: Public Discourse in the Age of Show Business, Neil Postman contrasted the dystopias of George Orwell’s 1984 and of Aldous Huxley’s Brave New World.
In the latter, a pleasure seeking populace enjoys free love, drugs, and constant entertainment, while making no meaning of any kind.
What Orwell feared were those who would ban books. What Huxley feared was that there would be no reason to ban a book, for there would be no one who wanted to read one. Orwell feared those who would deprive us of information. Huxley feared those who would give us so much that we would be reduced to passivity and egoism. Orwell feared that the truth would be concealed from us. Huxley feared the truth would be drowned in a sea of irrelevance. Orwell feared we would become a captive culture. Huxley feared we would become a trivial culture, preoccupied with some equivalent of the feelies, the orgy porgy, and the centrifugal bumblepuppy. As Huxley remarked in Brave New World Revisited, the civil libertarians and rationalists who are ever on the alert to oppose tyranny “failed to take into account man’s almost infinite appetite for distractions.” In 1984, Orwell added, people are controlled by inflicting pain. In Brave New World, they are controlled by inflicting pleasure. In short, Orwell feared that what we fear will ruin us. Huxley feared that what we desire will ruin us.
In Amusing Ourselves to Death – a book that has become only more relevant with age – Postman famously objects to the “And now, this” of television news. That is, he thinks it makes us worse when we watch a fifty second news segment about an ongoing genocide, followed by the news anchor saying “And now, this” and showing us footage of a prize-winning pumpkin from the county fair.
Postman died in 2003 and could scarcely have imagined that we would consume “content” in this manner on average, four to five hours per day, including (or especially!) on the toilet.
I was a standup comedian in the early 2000s when the iPod (for the young people: like an iPhone, but it only plays music, has no touchscreen, and only connects to your computer with a cable) came out.
Four years after the iPod came the iPod Shuffle, a smaller and simpler version perfect for clipping to your clothes while running. In exchange for its small size, users sacrificed control over what was played – different versions of the device held between 120 and 1000 songs, and the device simply played them randomly (hence the name).
I had a comedy bit about the Shuffle’s unhinged behavior, playing unrelated songs from different genres back to back! I acted it out on stage a bit: You better lose yourself in the music, the moment, you own it, you better never let it go, followed by Makin’ my way downtown, walkin’ fast, faces pass and I’m homebound. Funny, right? Those two songs would never be on the same album!
I’m not saying this was hilarious, and surely my Eminem impression was painful to behold, but people certainly understood that a joke was being made. Perhaps it failed as observational humor, but it received nods for being, at least, an observation.
This “joke” would not, today, even be recognizable as a joke, because the transition from a harsh song to a sweet song does not register to anyone as abrupt; we experience more abrupt transitions every few seconds every time we open TikTok.
YouTube is no better: you go on there to find a tutorial about how to replace the drum in your laser printer, you are offered a video about botched plastic surgery, and two hours later you are watching “WHAT IS THAT! Dr Lee Pops Satisfying Back Cyst | Dr Pimple Popper Reacts.” This may be unwanted, but the juxtaposition is not amusing to anyone; humor depends on an element of surprise and we are no longer surprised even a little bit by our worst impulses following us around the internet and snatching our attention wherever possible.
Neil Postman could blame television news stations for creating this type of content: there was a producer who decided the length and order of the segments, and there were anchors who assented to speaking to their audiences in this way.
Today, the algorithm produces such a “show” for each of us on demand. There is no attempt at coherence or any other over-arching quality whatsoever. There is no producer to blame. We can blame ourselves, but it seems like maybe the world’s fifth-largest company should also be on the hook a little bit.
Is a Virtuous Algorithm Possible?
We all know that a virtuous algorithm is entirely possible, but would be a horrible business move. What is this, an app for nerds?
I must admit, the closest technology in my own life to a virtuous algorithm is Amazon Fresh’s “Repeat Items,” which is an algorithm insofar as it invites me to add certain groceries to a repeat items list, the items on which then automatically appear in my cart every time I – somewhat dystopianly – buy groceries from the world’s third-largest corporation:
There are apps out there that promise a stream of philosophy quotes, or Bible study, or flashcards for aspiring veterinarians, and we all know that people download these apps with the best of intentions, but most of these same people spend hours every night watching what YouTube or Tiktok knows they really want.
Of course, most of the wholesome apps do not have an algorithm, or not a sophisticated one. A truly virtuous algorithm would collect data on what you impulsively click on, and serve you more edifying and helpful versions of it, becoming more and more edifying and helpful until you lose interest, then backing off a little. Could I be nudged from “Teen Killer CRIES at 25 year sentence” to “Improvements in DNA Technology Help Find Missing Teen!” to “Solving Crime with Calculus: Let’s Practice Implicit Differentiation”?
During the early 2025 discourse about whether Trump would or should ban TikTok in the US, part of the debate was about what effect the app has on young people; it came to light that China gives our young people an endless stream of addictive videos, but gives their own youth an app called Douyin, which requires face and age verification and limits users under 14 to forty minutes a day of educational content, health videos, museum tours, and science experiments.6 These videos are individually vetted, and I don’t know if the algorithm induces users to greater virtue over time, but this is likely the app closest to doing so. Surely the engineers behind it have the ability.
Inducing children to use this app and not a more exciting alternative, of course, requires a level of state control few Americans would accept – for example, China also enforces time limits for online gaming by maintaining a central database of gamers using facial recognition.7 Your virtue outside of a command economy is, of course, your own business.
New Problems Are Old Problems
About 300 years after Aristotle’s death, Seneca had an experience (recounted by me, here) very similar to my Youtube crime-watching spree: he decided on his lunch hour to treat himself, take a little break, go see a show at the arena!
But the man lives in ancient Rome, so he watches two guys with no armor stab each other with spears til one of them dies, then the winner has to have a stabbing fight with the next guy, etc., until the winner gets fed to the lions, and during halftime the crowd starts shouting for some throats to get cut because they haven’t seen enough blood yet. And Seneca is like, can’t a man just relax around here? He comments that he feels defiled after the experience: Rome will do that to you.
Lest his friend Lucilius respond (the men communicated by letter, responses took awhile) that there is nothing wrong with a criminal receiving a just punishment, Seneca wrote8:
You may retort: “But he was a highway robber; he killed a man!” And what of it? Granted that, as a murderer, he deserved this punishment, what crime have you committed, poor fellow, that you should deserve to sit and see this show?
Perhaps this is the moral of the story: Granted, a doctor may be popping pimples, and some poor soul must suffer from these pimples, but what have you done that you should deserve to sit and see this video?
You may want to watch the video. But do you want to watch the thousands of videos that will be shoved in your face as a result? And do you want to want them?
Endnotes
- Socrates’ views on pleasure appear in Plato’s The Republic, in which, sure, Socrates comes off as a real square who thinks your pleasures are inferior (to be fair, the Athenians did execute the guy). He argues that the just person is happier than the unjust person, and that virtue is actually more pleasurable than, you know, pleasure. But he makes an argument – one he actually dedicates to Zeus, as an Athenian apparently might do with a really important argument – that intersects intriguingly with our modern idea of the “hedonic treadmill,” introduced by P. Brickman and D.T. Campbell in a 1971 paper entitled “Hedonic Relativism and Planning the Good Society,” asserting that short-term spikes in happiness wear off and we all return to our baseline level of happiness, with the happiness-inducing stimulus no longer stimulating us once we adapt to it. Socrates’ observation is that most pleasures are “relative pleasures,” in that they are pleasurable in relation to previously experienced pain, and thus we adapt to them over time – but that pleasures that concern the higher parts of the soul are not relative, because they are not bound to the reality of physical things, and thus they do not diminish over time. These higher pleasures are, of course, those we can achieve through philosophy: a greater understanding of reality. A cynic might reply “Of course, a philosopher would be inclined to believe that the philosophical pleasures are the greatest and never decline with time, just like a soccer player might think the pleasure of soccer is the greatest and never declines with time.” (Start reading at the bottom of page 582, if you are so inclined!)
↩︎ - https://plato.stanford.edu/entries/aristotle-ethics/#EthiVirtDisp ↩︎
- Here seems like a nice place to toss out a link to Cindy Gallop’s longtime project MakeLoveNotPorn, which I think could be helpful to someone whose porn habits have damaged their real-life relationships (and which is ethically produced and queer friendly and all that jazz) ↩︎
- https://www.mozillafoundation.org/en/blog/mozilla-investigation-youtubes-dislike-button-other-user-controls-largely-fail-to-stop-unwanted-recommendations/ ↩︎
- My interest in this topic was further stoked when my husband remarked that YouTube was constantly trying to “redpill” him. Others on Reddit report the same. You see, YouTube’s recommendations are not based only on your watch history, but on the histories of other people who watch the things you watch (“Video Co-Viewing”). Thus, if you are interested in 9/11 truther content, for example, you are probably interested in beef tallow and other seed oil alternatives. My husband is interested in intermittent fasting and life extension; other people interested in these topics are also interested in “40 Minutes of Entitled Women Being Put in Their Place | Compilation.” It is truly unfortunate that the people attempting to live the longest are these ones. YouTube continues to show misogynistic propaganda to my husband no matter how many times he clicks “Not Interested.” YouTube surely knows he and I log in from the same IP address; it knows I live here. When the algorithm feeds my husband tradwife content, it knows I’m the wife! It is as though the algorithm is trying to oppress me, specifically. ↩︎
- https://www.bbc.com/news/technology-58625934 ↩︎
- https://www.technologyreview.com/2023/03/08/1069527/china-tiktok-douyin-teens-privacy/ ↩︎
- Seneca continues, “Even Socrates, Cato, and Laelius might have been shaken in their moral strength by a crowd that was unlike them; so true it is that none of us, no matter how much he cultivates his abilities, can withstand the shock of faults that approach, as it were, with so great a retinue. Much harm is done by a single case of indulgence or greed; the familiar friend, if he be luxurious, weakens and softens us imperceptibly; the neighbour, if he be rich, rouses our covetousness; the companion, if he be slanderous, rubs off some of his rust upon us, even though we be spotless and sincere. What then do you think the effect will be on character, when the world at large assaults it!” https://en.wikisource.org/wiki/Moral_letters_to_Lucilius/Letter_7 ↩︎
Want to support GetBullish? Send a gift card to a friend or buy one for yourself that you can spend later on your holiday gift shopping.
Want to break out of the algo? Join the No Scroll September Challenge.
