Reinforcement and Convenience in Technology

When it comes to social media and technology, there’s not much that hasn’t already been said. Still, I want to give some pushback on some places here where I think that society as a whole is insufficiently worried.

The following analysis is a little muddled. There’s places where I conflate smartphones with the Internet with social media. I ask that you bear with me and squint at places where things might be unclear to look at the general shape of what I’m describing.

The two main points here that I’ll be circling around focus on reinforcement and convenience.

By reinforcement, I’m pointing to things which give you dopamine, which engage your brain’s reward prediction system. Actions that are reinforced get repeated. In several studies, this seems linked to novelty and temporal distance. EX: Getting a jelly bean immediately in response to answering a question correctly provides more reinforcement than receiving a promise of two jelly beans in a month.

By convenience, I’m pointing to how readily something is available, both physically and mentally (in the sense of affordances). Actions which are convenient are ones which both spring to mind easily and are relatively easier to complete. EX: It is more convenient to eat a sandwich on a plate next to your laptop than to head outside and order one from the store two blocks down.



Optimizing for the Wrong Thing:

Say you’re on Facebook. Maybe you’re looking to see what some old friends have been up to. Maybe you’re sharing an article you found quite informative. Say Facebook sees you’re on their site. As you’re their commodity, it’s in their interests to not exactly abide by your interests.

You want to get the info you care about and be on your way. Facebook wants to keep you staying and coming back as long as possible. This means it might consider things like showing you relevant content only once every three posts (to keep you scrolling, my dear) viable strategies.

Remember that Facebook is not your friend. It’s Out To Get You.

This isn’t exactly a case of Goodhart’s Law. It’s not that we’ve got an imperfect metric for “stuff people want on Facebook” that Facebook is optimizing the heck out of. It actually just cares about something different than what you want.

It’s like the self-help thing all over again.

But worse.



As a consequence of the above, one thing certain services might try to burden you with is notifications.

I think notifications are evil.

I think notifications are evil because they are real-time and you don’t know when they will appear. Part of this is because most applications notify you when other people digitally interact with you, which is dependent on them. This unpredictability, though, means that you’re a lot more driven to consistently check your profile for updates.

This sort of variable interval reinforcement schedule leads to a consistent rate of response—you know at some point updates will roll in, but not knowing exactly when means the thought of, “Hey maybe I got some new responses!” always keeps the possibility open.

Contrast this to the mail delivery schedule, which happens at a fixed interval. It’s far easier to predict and thus means you spend less additional cycles wondering if something new has come. There’s less compulsion to check your mailbox at every free chance you get.

Incidentally, this sort of scheme solves the Fading Novelty problem, but it’s tailored at resolving the novelty for the wrong sort of thing.



Separate Forms and Functions:

I’m a proponent of separating out things intended for different uses.

EX: Sleep researchers recommend that you should only sleep in your bed and do everything else outside of it.

Much the same way, I find smartphone problematic because of the fact that they’re multi-functional. I don’t think it’s a good idea to have easy contact to games / things which provide easy reinforcement on the same device as something designed for research / work.

For a quick intuition about why I think this is the case: your brain is always doing lots of associations between your internal state and external stimuli. Thus, having a similar set of pathways which are designed for quite different goals could lead to confusion and slip-ups. And given our predilection to slide down reward gradients, I’d predict that the slip-ups occur in the “play shows up during work” direction far more than the other one.


Out-competing Other Options:

Smartphones boot up faster than laptops. They’re more accessible. Sure, the difference isn’t very great, but trivial inconveniences hold us back more than you might expect.

Along the lines of deep work, I’m worried about things which provide all-too-easy access to reinforcement. When work that’s more meaningful delivers reinforcement at a slower pace, it loses out to other options.

I think it’s important to renormalize your scales for reinforcement, in a similar manner as precommitment—if you know that certain options the current you values are going to be devalued as a result of something akin a hijacking, then don’t board that plane in the first place.

The availability of such options weigh on your mind, whether you want them to or not.

I think that that little space of your brain that’s been occupied, saying things like, “Why don’t you come and visit me again?” is a drain on being able to focus. If you want to be able to do good amounts of thinking, having lots of additional responsibilities means there’s more weighing on your mind, meaning less space for yourself.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s