We are surrounded by defaults, like dominoes
Notifications arrive unless we change them. Feeds keep scrolling. Videos start playing. The next video is already waiting to autostart.
Defaults are useful. But useful doesn’t mean right for everyone.
Autoplay is a good example.
Without autoplay, a video ends and there is a small decision:
Do I want to watch something else?
With auto play, that decision point changes.
The next video starts automatically. You can still stop it, of course. But continuing now requires no action; stopping does.
That may sound like a small difference. Autoplay changes the experience
Autoplay isn’t simply a trick designed to keep you watching.
There are genuine benefits to it. It can be convenient, help you discover something you wouldn’t have searched for, and make passive entertainment easier. If you deliberately want to watch several related videos, not having to press play each time can simply make the experience smoother.
But convenience isn’t the only reason continuing can be appealing. Curiosity can pull us toward what comes next.
Research published in Proceedings of the National Academy of Sciences found that higher curiosity promoted continued engagement with unfolding information, even when people had the option to resolve their uncertainty immediately. In other words, curiosity can make the process of discovering what comes next rewarding in itself. [1]
So when the next video appears, the question isn’t always simply:
“Do I want another video?”
It can also be:
“What is this? Maybe the next one will be interesting.” Autoplay makes satisfying that curiosity almost effortless.
But there is a trade-off
Autoplay makes continuing easier. But that convenience also changes how the decision to continue is made.
You still have control. You can stop the video, turn autoplay off, or simply leave the platform.
But continuing and stopping don’t necessarily feel like equal choices.
Continuing keeps the possibility of an immediate reward available. Another interesting idea, entertaining moment, surprising piece of information, or satisfying resolution could be just there effortless.
Stopping means giving up that possibility.
The benefits of stopping can be real — more time, less stimulation, or keep attention budget for something else, but they may be less immediate.
Research on intentional stopping provides an interesting parallel. In an experiment examining choices to continue or stop an ongoing action, the presence of reward incentives led to fewer intentional stop choices. The study wasn’t about video or autoplay, but it suggests that rewards can make approach and continuation harder to interrupt. [2]
This creates an interesting asymmetry:
Continuing offers a possible reward now. Stopping gives up that possibility for benefits that come later
Autoplay doesn’t create the reward. But it can make the path toward continuation and satisfy curiosity almost immediately.
Feeling in control isn’t the same as actively choosing
Research on autoplay itself adds another layer.
In a study of 394 people, autoplay that could be switched off was preferred and produced a stronger sense of control than fully passive autoplay. Yet, compared with manually starting each video, it also increased inattentiveness to recommended content. Both perceived control and inattentiveness were associated with a greater perception of having gone down a “rabbit hole.” [3]
This is an interesting paradox.
You can feel in control while paying less attention to what you are choosing.
That doesn’t mean autoplay takes away your agency. It means the environment around the decision has changed.
What does autoplay cost?
A little extra from your attention budget.
A 2025 experiment with 76 Netflix users provides a useful illustration. When researchers disabled autoplay, participants watched about 21 fewer minutes of Netflix per day on average, and their average viewing sessions were about 17 minutes shorter.
We can’t generalize this as it was one controlled experiment with Netflix users. But it demonstrates something important:
A small change in the effort required to continue can change how much time people spend consuming content.
And that brings us back to defaults.
Autoplay doesn’t decide for you.
It changes which action is effortless.
So, should you turn autoplay off?
That’s up to you. But it’s better to make that decision consciously than to let the next video decide for you.
If autoplay makes your experience better, helps you discover useful content, or simply removes an unnecessary interruption, there may be a good reason to keep it.
The point isn’t that autoplay is bad, it is to notice what the default is doing.
Autoplay doesn’t take away your ability to choose. It changes the environment in which that choice happens.
You don’t need to make it a permanent rule.
Try turning it off for a week.
See whether you watch differently.
Then decide whether the convenience is worth the attention you spend on it.
Pause for a minute.
Let the thought settle.
You don’t need another article.
Take it with you and go back to your life.
References
[1] Hsiung, A., Poh, J.-H., Huettel, S. A., & Adcock, R. A. (2023). Curiosity evolves as information unfolds. Proceedings of the National Academy of Sciences, 120(43), e2301974120. https://doi.org/10.1073/pnas.2301974120 PubMed
[2] Weidacker, K., Kvamme, T. L., Whiteford, S., Valle Guzman, N., & Voon, V. (2021). Incentives and voluntary stopping: The intentional hand task. Cognition, 206, 104504. https://doi.org/10.1016/j.cognition.2020.104504 ScienceDirect
[3] Chen, C., Kang, J., Sajjadi, P., & Sundar, S. S. (2024). Preventing users from going down rabbit holes of extreme video content: A study of the role played by different modes of autoplay. International Journal of Human-Computer Studies, 190, 103303. https://doi.org/10.1016/j.ijhcs.2024.103303 ScienceDirect
[4] Schaffner, B., Ulloa, Y., Sahni, R., Li, J., Cohen, A. K., Messier, N., Gao, L., & Chetty, M. (2025). An Experimental Study Of Netflix Use and the Effects of Autoplay on Watching Behaviors. Proceedings of the ACM on Human-Computer Interaction, 9(2), Article CSCW030. https://doi.org/10.1145/3710928