Ironic anti-Beeminder argument

[dreev]

I often talk about Beeminder as the nuclear option. That you should beemind some bare minimum that you’d definitely be upset with yourself for not following through on, and then find other forms of motivation to make Beeminder moot. But still have the Beeminder graph to catch you if those other forms of motivation fail. (And for tracking/visualizing the progress in any case, if there’s inherent value in that.)

I’ve actually heard the amusingly ironic argument that maybe you shouldn’t have Beeminder as backup in that sense because the lack of a backup can serve as a commitment device to ensure that those other forms of motivation don’t fail. But I figure if that reasoning is appealing then you’ll probably love Beeminder as a primary source of motivation. : )

[philip]

I’ve heard that the presence of safety buffer and alarms causes operators to work the machinery harder than they did previously.

Without the alarm, an experienced operator has to gauge the machine’s tolerances manually, and tends to err on the side of safety. With the alarm, you can push performance until it’s constantly running in beemergency mode. That’s hard on the machinery, and hard on you (if we’re talking goals rather than machines).

Experienced operators of Beeminder get better at slope-setting so that not too many of our goals are in beemergency mode at any one time.

[drtall]

A rambly email that I sent to dreev before this thread existed:

Part I, WHEREIN I RAMBLE ABOUT HIERARCHICAL MOTIVATIONAL SCHEMES

Thinking about backup motivation systems gives me the same existential dread that I feel whenever I think about what it would be like if the afterlife is that there isn’t one and you are simply annihilated.

It’s like in Kindergarten when the teacher writes the naughty children’s names on the chalkboard. To some children that means absolutely nothing and to others (me) it keeps you up at night with anxiety. You might seek to fix that anxiety, and internalize the idea that your name on the chalkboard is meaningless. That might seem rational. But, at the end of day, all the children either end up behaving or expelled, so you might as well just go with the chalkboard name as your motivator and avoid detention, boarding school, a life of crime, etc. etc.

In the same way, it is important to not think too hard about how little Beeminder can actually, truly sting. It can take my money, okay. But I have enough money to pay. Or I could lie to Chelsea. Or I could dispute the charges with the credit card company, burn down the Beeminder office, etc. It’s a spiral of doom that is only avoided by keeping the derail as the fundamental unit of shame and anxiety.

To be clear I am only talking about commitment systems. I do get a lot of motivation from seeing green Beeminder graphs or Seinfeld chains within my goals. But this is all a fluffy type of self-actualization and not a commitment. Doing better than my commitment is not something I can commit to, by definition.

Part II, WHEREIN I CLAIM YOU CANNOT BEAT BEEMINDER

The idea of “backup” motivational schemes also feels like a contradiction from the perspective of want-can-will. Suppose you believe want-can-will, and you’ve already tuned all your Beeminder goals to the maximum you can do given want-can-will.

Now you want to take on an additional commitment that’s even harder than your Beeminder goals? You are doomed to fail by the precepts of want-can-will. Maybe you’ll do okay for a while, but it isn’t sustainable. If it were sustainable to do more, then your Beeminder goal was set incorrectly.

Now, is it possible that you have some kind of fuzzy motivational system that increases the maximum you can do? Maybe! If you find another tool is letting you become less Akratic and able to do more under want-can-will, that is a great discovery. However, this now means your Beeminder goal is set incorrectly. If you don’t reset your Beeminder goals for your improved want-can-will, then you’re always at risk of discarding this new thing and reverting to your old ways.

Beeminder isn’t for everyone. There’s people who will do better on other systems and that’s fine. But I don’t believe that anybody can benefit from combining Beeminder with another commitment system if they are tuning their goals according to want-can-will.

[drtall]

@philip

That’s hard on the machinery, and hard on you

But in the metaphor, the work of

@philip

has to gauge the machine’s tolerances

is also falling on you. I’d rather be dealing with a pre-computed boolean “Do I need to do X today?” than dealing with the judgement call / thinking problem.

[insti]

I don’t think it’s about it being the nuclear option, or the barrier of last resort, I CAN pay my taxes without using Beeminder, the threat of being put in jail makes sure of that.

In many years of looking I’ve not found another motivational option that works.
The process tended to go like this:

6 months of putting it off + 1 month of seriously putting it off + 1 week of ‘I really must do that’ + 3 days of ‘I really need to get this finished and submitted today’ + 12 hours of ‘I hate my life why didn’t I do this earlier’ + 4 hours of ‘damn the tax website is slow today’ + 30min of ‘wow that was easy, I should have done this 7 months ago’ + 2 days of ‘Ok, next year I won’t leave it to the last minute’

Compare this to the Beeminder way:

“I just need to do 7min work on my tax today” or I pay Beeminder $5.
(I end up doing 25 and am able to forget about it completely for another week knowing that Beeminder will remind me again when it’s time to do more work.)

And for other things that no-one else but me cares about, they just don’t get done without Beeminder.

I still could not use Beeminder, but it makes my life a lot more pleasant, and leaves me with fewer regrets.

[insti]

My Beeminder irony is that I now procrastinate by doing Beeminder related things like posting on the forum.

[dreev]

@drtall

burn down the Beeminder office, etc

Well that escalated quickly!

(I’m kidding; don’t take any of this out of context, future readers! It’s a pretty profound point about setting up the mental framing where the yellow brick road is the fundamental motivator.)

I’m still mulling your counterargument in Part II, which I like but of course I’m too motivated to like (so many layers of irony) for me to trust my assessment of it yet.

PS: Let me also link to want-can-will for any newbees not in the know.

[tracy_reader]

But the people setting the alarms and safety buffers will have an incentive to err on the side of safety, as it’s a bit embarrassing if the machine blows up before the alarm goes off.
And it’s costly to build a machine with capabilities that never get used (to put it in environmental terms, every tonne of steel we don’t have to smelt or recycle is a good thing), so if the operator does run the machine harder while still within the safety margin, then that’s a good thing.

To apply the analogy to Beeminding, I kinda thought the point was to be hard on me so as I could be more productive in the long run.

[philip]

@tracy_reader

But the people setting the alarms and safety buffers will have an incentive to err on the side of safety, as it’s a bit embarrassing if the machine blows up before the alarm goes off.

My understanding is that this leads to a cascade of safety alarms, which experienced operators learn to ignore…

So maybe it’s the nature of the alert. If the alarm says that the machine’s about to blow, it’d better be true. Kind of like the ‘your flight is leaving’ announcements in airports, at least in the UK, which used to be almost-true and now seem to mean that the plane has arrived at the gate and you should think about heading over.

@tracy_reader

To apply the analogy to Beeminding, I kinda thought the point was to be hard on me so as I could be more productive in the long run.

Absolutely yes. Sustainably hard on myself, so that I get more of what I want done, done, and I am still beeminding in the long run. As experienced operators of Beeminder, the thing we’re avoiding burning out is ourselves.

[philip]

@philip

That’s hard on the machinery, and hard on you

@drtall

But in the metaphor, the work of “has to gauge the machine’s tolerances” is also falling on you. I’d rather be dealing with a pre-computed boolean “Do I need to do X today?” than dealing with the judgement call / thinking problem.

It’s turtles all the way down, I’m afraid. You’re in charge. Not even the mighty Beeminder can absolve you of that.

In the moment of doing, ok, you need clear indicators of what to do now/next/today.

When you have the luxury of time, you also think about what you want to accomplish, how you might go about that, and what actions you can actually and usefully take.

That’s one self, from a hopefully better/longer perspective, deliberately constraining the options of a slightly-future self. At another level of recursion, there’s another self thinking about how to stop the one self from burning out the slightly-future self by beeminding all the things, “all at once and much, much too completely”.

By reducing things to near-binary choices, we’re surrendering agency to our past selves through the mechanism of Beeminder.

In the same spirit, one self sets the alarm clock, imposing on a slightly-future self the binary choice of getting up or hitting snooze. It often works better if you do what your past-self thought would have the best result over time, at least within relatively short time horizons in which life almost always remains stable.

Aside: my friend Tom Graves has a lovely model called SCAN that includes time horizons and the impact on complexity. One of which is that in the right-now time, everything needs to be binary. Here’s a presentation on SCAN that he gave (to a military audience) earlier this year.