The question usually arrives in a self-accusing form. Is this a real thing, or am I just weak?
It is a fair question and it has a better answer than most people expect, because the mechanism is specific, it is documented, and a fair amount of it was built deliberately.
Is phone addiction a real addiction?
Clinicians argue about the word, and that argument is not settled. "Addiction" carries diagnostic weight that nobody should hand out casually, and the research literature mostly uses narrower terms.
What is not in dispute is the machinery underneath. The reward system that makes the feed hard to leave is the same one studied in every other compulsive behaviour, and the way feeds are built engages it with unusual precision.
So the useful question is not whether it qualifies for a label. It is why it works on you, and whether that mechanism gives you anything to act on.
Why is the phone so hard to put down?
Because dopamine does not do what almost everyone thinks it does.
Wolfram Schultz's work established that dopamine does not encode pleasure. It encodes prediction error: better or worse than expected. And it migrates backwards onto the earliest reliable predictor of a reward.
Applied to a feed, that means the charge does not live in the video you enjoyed. It lives in the swipe, in the half second before you know what is coming. The anticipation is the event.
This is why the experience is so consistently reported as joyless. You are not there for what you find. You are there for the moment before you find it, several hundred times an evening.
Source: Schultz W, "Dopamine reward prediction error coding," Dialogues in Clinical Neuroscience (2016).
Why are feeds designed the way they are?
Because there is a mathematically optimal level of uncertainty for producing pull, and an infinite feed sits on it.
Fiorillo, Tobler and Schultz recorded dopamine neurons under different reward probabilities and found the response ramps hardest at roughly 50/50 odds. Not when a reward is certain. Not when it is unlikely. When it genuinely might or might not arrive.
A feed where the next item might be extraordinary and might be nothing is tuned to precisely that point. It is the same principle a slot machine runs on, and it is not an accident of the medium.
The behaviour-design field that produced these products is open about its methods. Persuasive technology has been taught, published and sold as a discipline for two decades, and the trigger-action-variable-reward-investment loop was written down as a manual for building exactly this.
Why doesn't willpower work against it?
Because wanting and liking are separate systems, and only one of them is driving.
Berridge and Robinson's incentive-sensitization work separates the circuit that drives pursuit from the circuit that produces enjoyment. Repeated heavy use sensitises the first and leaves the second flat. You end up craving something you are not enjoying, which is precisely what people describe.
It also explains the timing problem. The reach fires before the thought to scroll arrives, so any plan that depends on catching yourself in time is depending on a moment that has already gone.
Can it be reversed?
Yes, and faster than the framing of "addiction" tends to suggest.
In February 2025, Castelo and colleagues published a randomised controlled trial in PNAS Nexus: 467 participants, mobile internet blocked for two weeks. They reported improvements in mental health, subjective wellbeing, and objectively measured sustained attention.
Two weeks of reduced input produced measurable change. The mechanism that built the problem also unwinds when the input stops.
Source: Castelo et al., PNAS Nexus, February 2025.
What do you do with this?
The useful consequence of all this is narrow but real: stop aiming at the decision, and start aiming at the reach.
- Put the friction at the door. A block that fires without consulting you is not there to be strong. It is there to interrupt the loop at the only point where interruption still works.
- Choose in daylight. Pick the apps when you are calm. Anything left to the evening is decided by the version of you with the least resistance.
- Expect the flatness. When the input drops, ordinary life reads as boring for a few days. That is the baseline recovering, not evidence that this is not working.
- Watch one urge all the way through. It rises, peaks and drops when nothing feeds it. Almost nobody has seen that happen, which is why it sounds like a claim rather than a description.
None of this needs an app. Block the apps in your phone's settings, and when the gap opens, sit in it without reaching.
What an app buys you is the part you have already proven you cannot do alone. Boredom's Daily Reset is three guided minutes placed inside the gap the block opens, so there is a voice in your ear for the stretch where you would otherwise reach. What fills the space afterwards is yours, and it is analogue and free.
The short version
The pull is real, the mechanism is documented, and a great deal of it was engineered on purpose against a reward system that was never built to withstand it.
That is a design flaw in the environment, not a defect in you. Nobody clears it on your behalf, but you are not fighting a character weakness, and the thing you are actually up against is only about a minute long.