Personal automation for augmenting intelligence: A shortcut to stay on task
# Personal automation for augmenting intelligence: Monitoring, Context-setting, and Generating
I was recently interviewed on an episode of Automators. Connecting with Rosemary and David was awesome. We chatted about Obsidian, how some automation tools have made parenting a little easier, and a topic near and dear to my heart: automation for augmenting intelligence.
Augmenting intelligence with technology—especially with computers—is an old idea. It was popularized by the field of cybernetics and early computer researchers such as Doug Engelbart.
In my view, the basic idea of augmented intelligence is simple. Computers are smart in some ways and dumb in others. Humans are dumb in some ways and smart in others. If we can pair the two well, we should be able to be smart, together, most of the time.
For a long time, however, approaches to augmented intelligence have been kinda inaccessible. For example, the US Defense Advanced Research Projects Agency (DARPA) Augmented Cognition (AugCog) project explored using complicated and costly technology such as Electroencephalogram (EEG) readings to monitor the state of mind of warfighters and to adjust the software they were using in tandem.
DARPA’s AugCog work is wild, inventive, and clearly important to emergency situations like combat or surgery. For most of us, though, it’s an unimaginably impractical set up.
Yet we regularly use computers to augment our cognition. Any time we set a reminder or a timer, we’re offloading a thinking task onto the computers on our desks or in our pockets.
Personal automation tools extend this idea considerably. If you’ve used Shortcuts or a macro app of some variety (e.g., Keyboard Maestro), you’ve used personal automations. You have probably used these tools to do something more quickly or without error.
But what if we thought a little bigger? How might personal automation be used to amplify our intelligence beyond speeding up routinized workflows or limiting our mistakes? What would it look like if these tools helped us to monitor our state of mind, reframe problems, or think differently?
I think they can. I’m proposing a simple framework for “personal automation for augmented intelligence.” Consider how you might set up Shortcuts or macros to support monitoring your state of mind, changing your context to suit the current task, and generating new ways of thinking about whatever you’re doing.
- Monitoring: time-, location-, and other context-based triggers that check in with your current state of mind
- Context-setting: changing your digital environment to emphasize what’s important for your current task
- Generating: questions, challenges, and other prompts that reframe your thinking about whatever you’re engaged in
These might sound fancy, but they can be designed with a few simple building blocks: appand window-switching and scheduled notifications (via reminders).
In the rest of this post, I’ll walk you through a set of shortcuts I put together using these ideas. The shortcut helped me face off against anxiety and get through my PhD’s comprehensive exams.
# An example augmented intelligence shortcut
This is a simple demo of the use of personal automation for augmented intelligence.
The shortcut itself is pretty idiosyncratic. I would be happy if you found it useful immediately, but I doubt that will be the case for most people! Instead, I hope you can take some ideas from it to adopt into your own ways of thinking and working.
# Why I built the Mise en place shortcut
Most PhD students have to write comprehensive exams (aka “comps”) before they can begin the research work that will make up their dissertation.
A comprehensive exam is aptly titled. The idea is to test a student’s knowledge of their entire field of study. So, most comprehensive exams require that a student deeply understand dozens or hundreds of papers… and usually a few textbooks to boot.
I have always been a bad procrastinator. I think my procrastination habits are rooted in anxiety.1 As you might guess, comprehensive exams are very anxiety-provoking. So, when I first began preparing, I had a lot of trouble. I tried to draw on best practices—breaking down the task, defining the end goal, practicing mindfulness—but I was struggling.
One of the things I would use to procrastinate was Shortcuts.2 Then, one day, I realized: why not build an automation that guides me through these best practices to conquer procrastination?
That’s where Mise en place came from. The name is stolen from the French culinary concept of “setting in place” everything you need before you begin cooking.
The core concept of the Mise en place shortcut is simple: lead me through a preparatory ritual to reduce anxiety, and then add some check-ins to help catch me in the act if I go too far off-task.
# How it works
The whole thing is actually a set of nine shortcuts. Mise en place calls three others, each of which call more.
I designed it this way for reusability: some of these inner shortcuts I actually use in other shortcuts, too. By breaking them up (or decomposing them), I can use the parts in multiple places, and if I change or have to fix anything, I don’t need to implement the change in multiple places.
Below, I explain each of the individual shortcuts.
# Mise En place
This is the parent shortcut. It’s the only one you’ll use in practice (unless you reuse the others elsewhere, as I describe above).
Mise en place does the following:
- It asks how long you plan on working for (e.g., for your current “block” of work).
- It turns on a focus mode (née Do Not Disturb) for the duration of that work.
- It asks what you’ll be working on during that time.
- It then opens Pause, a meditation app, so that you can clear your head/be mindful before diving in.
- When you return from Pause, it runs Focus sounds, described below.
- It then runs Create check-ins, described below.
- Next, it prompts you with a reflective question, asking you to visualize what success in this block of work looks like.
- Then it runs the Visualize shortcut, described below, in order to anchor the work you’re going to put in into some vision of the future you’re working towards.
- Finally, it schedules a notification (via reminders) that celebrates the end of the block when the duration has elapsed. (That reminder includes a link to the Visualize shortcut, if you want a reminder of what you’re now one step closer to.)
# Configuring Mise en place
When you install it, choose the focus mode that you’ll enter after launching the shortcut, or just Do Not Disturb.
If you do not have Pause, you can choose to launch another meditation/mindfulness app or just delete that action.
Later, you may want to customize any of the text for the shortcut’s notifications and reminders.
# Focus sounds
This is a pretty simple shortcut. It asks if you want to play something different from the phone. This is just in case you already had your groove on, you were planning on playing music or white noise from another device, or you were just appreciating the silence.
If you say yes, it’ll open myNoise, my favourite white noise generator. You can then launch any background noise you want.
When you return to Shortcuts, it’ll let you layer on more music. It gives you the choice of:
- A Focus playlist (from yours truly—6.2 days of no-vocals music) via a separate shortcut,
- Brain.fm, an ambient focus music generator,
- Endel, another ambient focus music generator, or
- just Background noise only.
If you select any of the first three options, it’ll play the playlist or launch the app.
# Configuring Focus sounds
Replace the actions with your desired background noise/music apps/playlists. I love having rain + something else, which is why this shortcut walks you through creating two layers.
# Focus playlist
This thing is super basic. It just plays my Focusing playlist on shuffle.
Again, I have this set up as a shortcut in case I ever change my main focus music playlist. If that ever happens, I fix just this shortcut, and the change will take root in all the shortcuts I use this shortcut in.
# Configuring Focus playlist
Choose a playlist of your own.
The basic concept is simple enough (although it may seem fluffy to you). First, create a list of concrete, tangible things you’re looking forward to or that you’re working towards.
In order to build up some momentum or inspiration before heading into something intimidating, this shortcut set will show you one of those items. Then, if you want to see more, you can keep going until you want to stop. (That’s the recursive part.)
There are three sub-shortcuts here: Get visualizations, Visualizing positive futures, and Recursive visualize.
# Get visualizations
Pretty simple: receive a text file, read that text file, and add every line that isn’t a heading to a list. That list is the “visions” that the visualizations subroutine will use.
# Recursive visualize
This is how we loop this visualization subroutine. Recursive visualize calls the Visualizing positive futures shortcut, shows the result, then asks if you want to continue. If so, it runs itself again.
# Visualizing Positive futures
Again, very simple: receive a list, pick a random item from that list. This one takes in the list from Get visualizations and passes on a random future you’re looking forward to.
# Configuring Visualize
- Make sure you have a text file that is a list of images of the future you’re looking forward to. (See the example below.) Configure the Get visualizations shortcut to grab that file.
# Create check-ins
Create check ins receives a duration (a number of minutes), and it uses that duration to set up a random number of random check-in notifications at random intervals.4
I could explain the shortcut in excruciating detail, but if you’d be willing to read that detail, you’re probably better off just looking through the shortcut to understand the logic.
After the shortcut decides when each check-in should be, it runs a helper shortcut called Unstuck - Simple to actually schedule the check-in notifications. That one’s explained immediately below.
# Configuring Create check-ins
No configuration needed!
If you want, you can adjust the number of intervals the shortcut creates by playing with the math in the shortcut.
# Unstuck - Simple
Wouldn’t it be nice if someone said “it’s okay,” or “maybe you should switch things up” when you get stuck at work? Well, that’s what this shortcut does.
Unstuck - Simple receives a date and time (from Create check-ins). Then, it creates a check-in prompt. It does this via simple ad-lib randomness. Each a prompt is constructed from four parts:
- transition, and
The shortcut chooses a random phrase from a list for each of those pieces. It puts them together into one string, then it schedules a reminder with that text for the received date and time.
# Configuring Unstuck - Simple
You simply need to choose a Reminders list. I used Gestimer, the list the macOS Gestimer app uses to put a reminder countdown in your menubar. But you can use anything you want.
# Setting it all up
- Download all of the Shortcuts at once. Don’t run or edit any of them until you’ve installed all of them. (The order you install them in shouldn’t matter, but let me know if it seems to…)
- Some shortcuts will ask you configuration questions as you install them. Follow the instructions for the relevant shortcut above to make sure you’ve set it up correctly.
- After you run each shortcut for the first time, it will probably ask your permission to do a bunch of things. That should only happen once.
# Links to the shortcuts
- Mise en place
- Focus sounds
- Focus playlist
- Get visualizations
- Recursive visualize
- Visualizing positive futures
- Create check-ins
- Unstuck - Simple
# That’s it
Well, I hope this gave you some ideas for how to use personal automation to augment your own intelligence!
If you have anything to share, feel free to reach out on Twitter.
A fear of failure (or something) drives me to avoid the task at hand, diverting instead to activities that alleviate the fear, if only for a moment. ↩︎
Playing with Shortcuts is sorta like solving puzzles for me. ↩︎
Recursion is when something invokes itself. In computer science, this is used to loop until some condition is met. In this case, the recursion stops when we tell it to stop. ↩︎
I’m using randomness as a substitute for more intelligent possibilities. Perhaps someday an app can use some kind of metadata or, yes, cognitive monitoring to identify the best times and ways of checking in with a user! ↩︎