My daughter was a late talker. At three we were suspecting autism. In my search, one of the centers offered me an expensive program to fix her with one condition: I should leave her the first week, and I shouldn’t be around. This of course raised all the red flags inside my head. I was torn between someone offering me a treatment and my instinct that this sounded wrong. That led me to find where all these programs are rooted. A study published in 1987 by Ole Ivar Lovaas on how to make autistic children “indistinguishable from their peers.” Not happier. Not more capable on their own terms. Indistinguishable.
The method is 40 hours per week of discrete trial training—a phrase that sounds clinical enough to slide past without realizing what it actually meant for the child to be sitting in the chair, rewarded with praise and food, punished with electric shocks when not in compliance—”normal earned punishment” per Lovaas’s words—with a single objective “Be like me”. The child won’t eat, won’t be praised unless they act like what’s so-called normal. The child’s natural behavior is destroyed and is replaced with a robot, a child forced to act to look normal to get basic stuff like food! That was nearly four decades ago. The electric shocks Lovaas used have been abandoned. No more food deprivation. ABA practitioners kept refining and now distance themselves from the specifics of his program, which is understandable given what was involved. But the architecture never changed: correction needs to be administered through externally imposed behavioral schedules. It never disappeared; actually it got expanded from the university clinic at UCLA to the App Store.
Most of the habit-tracking apps in 2026 follow that architecture to the book. Users are rewarded with ”streaks,” miss one day and the streak collapses to zero. Red replaces Green, Notifications worded as warnings for “accountability“. Features that allow others—a friend, a partner, a therapist—to monitor whether the user is in compliance or not. Miss a habit and the avatar’s health goes down; miss more and the avatar dies!
We have been using shiny wording like “self-improvement,” “building better habits,” and “personal growth”; however, the reality is “comply”. Most of these applications are behavioral compliance systems dressed as mobile applications. Their design logic descends from an intellectual tradition with a specific theory of what a human being is: an organism whose behavior can and should be shaped through positive and negative reinforcements. Any deviation isn’t treated as a signal, or an opinion or a reasonable response to unreasonable conditions. It is more of a failure. The organism failed.
The Skinner Box Has an App Store
This isn’t a new playbook. B.F. Skinner wasn’t the origin of the observation that consequences shape behavior, but he was the one who gave it a formal structure good enough to be industrialized. Make the reinforcements unpredictable and the result is an addict. Skinner himself identified gambling as the obvious human parallel.
Silicon Valley adopted the framework. Nir Eyal’s Hooked: How to Build Habit-Forming Products, published in 2014, laid out a four-step cycle—trigger, action, variable reward, investment—which follows Skinnerian reinforcement schedules to the letter. The book became an essential in any product team bookshelf. B.J. Fogg, who ran the Behavior Design Lab at Stanford, had been training product designers to exploit those concepts and treat user behavior as an engineering problem: increase motivation, reduce friction, calculate triggers (variable ones) and the desired behavior emerges like water finding a channel.
Habit-tracking apps added a new dimension that neither Eyal’s book nor Fogg’s lab addressed. Using operant conditioning to change user behavior from what is considered non-compliance. Unlike social media, where the goal is more engagement, or Duolingo where the goal is learning a new language, with mental health apps (like ADHD habit trackers), a streak miss means a failure. The app has already decided the user’s patterns of attention and motivation are a deficit that needs to be fixed, and the streak is just the enforcement tool.
What Compliance Looks Like When It’s Wearing a Friendly Interface
Autistic adults who went through these programs tell a different story. The say the programs are training them to override their own distress signals, to shut down the behaviors their bodies actually need for regulation. The European Council of Autistic People said in 2024 that ABA’s objective was never to accommodate difference. It was to eliminate it.
The word “compliance” is never used by habit-tracking apps. They use “consistency.” Or “accountability.” Or “building momentum.” But if you remove the dust, it screams compliance.
A streak counter is continuous reinforcement rendered in an app. The more you are in compliance the counter goes up, and the more it goes up the more emotional weight is built. The user is not being rewarded for learning something, or for feeling better, or for discovering which approaches work for their particular brain. They are being rewarded for showing up on schedule. The reward is just a number that goes up, and the punishment is a number that goes down. Nothing more than “compliance training.” What is being enforced is not insight or wellbeing but pretending to be normal. The structure is Skinnerian down to the wiring.
Losing something hurts twice as much as gaining the same thing feels good—that is Daniel Kahneman and Amos Tversky’s prospect theory, published in 1979. Mapped to streak, a person on day 47 doesn’t process day 48 as a mild reward but as a threat of losing the streak. The app has built a trap based on fear of loss. For people diagnosed with ADHD this is even worse. A study showed that 34-70% of the people with the diagnosis experience significant difficulty regulating emotional responses to perceived failure.
A 2016 paper presented at CHI, the Conference on Human Factors in Computing Systems, stated the obvious: people leave habit-tracking apps because of streak dependency. Not boredom, not forgetting. What was promoted as a fix ended up being the problem.
The Disease Model as Default Setting
The reason nobody notices the disease model in these apps is because it’s disguised as “helping.” The app tracks your morning routine, your meds, your focus time. It looks like a helpful checklist. But every item on that checklist is built around what your brain supposedly can’t do. The app doesn’t say “we believe your brain works differently.” It says “let us help you stay on track.” Same message, friendlier packaging.
Sami Timimi, a child and adolescent psychiatrist, has spent decades arguing that ADHD is better understood as a cultural construct than a neurobiological disease. He doesn’t say there is no struggle with attention, or that executive function difficulties don’t exist. The point is the diagnosis takes real struggles and locks them into a medical framework that blocks other explanations and serves institutional interests. Once you get the ADHD label, everything is filtered through it. Messy desk? Symptom. Unfinished project? Impairment. The label stops describing and starts prescribing how you see yourself. ADHD apps enforce this loop by tracking “symptoms” and rewarding “management.” It is not just reflecting the diagnosis—it’s hardening it. Every missed notification is a daily reminder: “I’m failing again.” Emaline Friedman called this out: these technologies redraw what counts as a “symptom” vs. what counts as normal.
Thomas Szasz made the structural version of this argument decades before anyone had built an ADHD tracker. Psychiatry’s core function, Szasz contended, was social control rather than medical treatment—the vocabulary of diagnosis and “therapy” lending an air of clinical legitimacy to what was, at bottom, coercion. The habit-tracking app is an oddly efficient heir to this tradition. No one is being institutionalized. No one is being forcibly medicated. But the person diagnosed with ADHD who opens the tracker each morning and confronts a red indicator where yesterday there was a green one is receiving the compressed enforced message: “your brain didn’t perform correctly, try again.” This is the message that has organized psychiatric treatment for a century and a half: the natural state of this mind is the problem, and the only acceptable response is correction.
Shame as a Design Feature
There are two ways to motivate someone. One is genuine interest (intrinsic motivation). The other is guilt and shame (introjected regulation—Deci & Ryan’s term). From the outside, both look the same—the person does the thing. Inside, the cost is completely different.
Habit-tracking apps operate on the latter. The streak does not make the habit interesting or infuse the task with meaning. That shame-based approach actually destroys the only thing that produces lasting behavior change: autonomy. In short: these apps use the one motivational strategy that psychology has proven kills long-term motivation.
Alfie Kohn put together decades of research in his book Punished by Rewards, he showed, confirmed across 128 studies, how rewards kill motivation. Grades, gold stars, streak counters—all these are damaging the reward system. Rewards do the most damage when the person already found the task interesting and you shift this into the reward. The reward replaces the interest and they stop enjoying it. The effect is compounded with individuals labeled with ADHD where many experience rejection sensitive dysphoria—an intense emotional pain from perceived failure. It is exceptionally harmful to the exact people it claims to help.
Accommodation Against Compliance
The neurodiversity movement—despite debates and criticisms—proposed a different approach. Instead of how to make individuals act normal, it asks “what does the environment need to change so this person can thrive?”
Instead of compliance where you change—schedule, penalty, failure record—we accommodate, and new questions start to arise. When does YOUR attention peak? What task order fits YOUR brain? What sequence of tasks align with YOUR rhythms? These are the questions that were suppressed by the ADHD label and the coercion to act like others.
It’s not rocket science to address this. Yet there are barely any attempts to do so. The habit-tracking market—estimated at over a billion dollars and still expanding, though evidence that these apps produce durable behavioral change remains thin—is flooded with apps following the same book. Set goals, monitor adherence, penalize deviation, reward consistency. The trap is here. Consistency is treated as the goal, but consistency is only a problem if inconsistency is a disease. And inconsistency is only a disease if you treat the neurotypical pattern as the biological default rather than just one way among many.
Why don’t the apps question any of this? Because they don’t have to. They have a working model on the surface, a model backed by science and a proven business model—Duolingo—that generates money. They feel good about what they are doing by pushing toward regularity and they get to call themselves therapeutic. The Skinner Box requires no justification if the organism inside it has already been classified as broken.
What Would It Mean to Build Something Different?
The question has never been whether people need support or not. It is “what kind of tools, built on which architecture, whose version of ‘better’”?
Is it a tool built for compliance that renders failure visible and enforces being different, a tool built on the foundation of Skinner’s reinforcement schedules and Lovaas’s discrete trial sessions? Where shame is the driver?
Or is it a tool built on accommodation—what you did today, how you feel, and how to adjust for a better tomorrow. A tool with no reset to zero. With no assumption that the user is broken.
The gap between the two philosophies is not cosmetic. It is political in the deepest sense. Is cognitive variation something to suppress or accommodate? Unfortunately the industry chose suppression. They built a system over Skinner’s pigeon experiments and Lovaas’s child conditioning just with better packaging. They stripped it of the cruelty, rendered it as perfect UX with soft gradients and rounded corners, and sold it as “self-care.”
It is true that nobody is forced to download an app, but it is the same choice I faced nine years ago. People are given a choice between what’s called a cure or finding another path. The choice isn’t really free, it is shaped by a framework that already decided this is the only way. Until the paradigm is challenged, the streak will continue counting, resetting and turning red.
The streak is not a feature. It is a philosophy. And the philosophy, when examined without the cushioning language of wellness and self-improvement, turns out to be the same one that has always powered the biomedical model of mental distress: the problem resides inside the individual, the solution is behavioral correction, and the only question that matters is whether the patient complied.
***
Mad in America hosts blogs by a diverse group of writers. These posts are designed to serve as a public forum for a discussion—broadly speaking—of psychiatry and its treatments. The opinions expressed are the writers’ own.