Wrestling with the Angel of Death

Clive Thompson and the dead-end of task management.

In Hundreds of Ways to Get S#!+ Done—and We Still Don’t, Clive Thompson finds a lot wrong with how we approach task/work management, but he simply accepts a lot of preconceptions without really examining them, I feel.

The most fundamental preconception -- never examined by Thompson at all -- is the binary nature of tasks in nearly all task management tools. They are open and active or closed and inactive. But this is an enormous oversimplification of how we actually do work. Tasks can have many states, not just two, something I elaborate on, later. I believe that a great deal of Thompson’s angst — and other users of task management tools — is related to this mistaken perception.

As I pull selections from Thompson's essay, I am going to intersperse comments, to examine his preconceptions.

Most common office tasks have well-settled software “solutions.” If I asked you to write a document, you’d probably use Word or Google Docs. To make a presentation, you’d pull up PowerPoint or Keynote or Google Slides.

Not so for to-dos. There is no Way That Everyone Does It. It’s a crazy Pokémon deck of options: Trello, Todoist, Gmail’s tasks, Microsoft To Do, Remember the Milk, Things, OmniFocus, Any.do, Evernote’s Tasks, and Clear, to name just a few. And that doesn’t even count the whackload of us using one big ol’ Notepad file on our computers, or even plain old paper.

Task management is one circle in a much larger Venn diagram of contending approaches to keep track of things, approaches with overlapping capabilities and very different structures. Task management is something many tools offer in very different ways.

Work management (task management's more enterprise-oriented big brother, with offerings like Asana, Trello, Monday and others) is under increasing competitive pressure from other approaches, such as spreadbases (like Airtable, Notion, Coda, and others), work processing (like Dropbox Paper, Salesforce Quip, Roam Research, Obsidian, Typora, and others), and business operating systems from Microsoft 365 and Google Workspace, which both offer task management as fundamental elements.

Thompson is right that there is no single common way to do task/work management. But the entire landscape is in flux, and not just because of flaws in task management tools, per se. But the basic model of task management -- binary task states, and lists of tasks, where tasks can have attributes like due dates, priorities, comments, and attachments -- has been replicated in thousands of apps. That model has been very influential, but no single task management tool has become totally dominant.

The creators of personal to-do apps—or task management software, as it’s sometimes called—generally agree that they haven’t cracked the nut. Every one of these apps attempts to handle the same kind of basic actions: Give people a way to write down tasks, like “Get milk” or “Finish the sales memo,” and offer tools to sort and prioritize those tasks. Ideally, that improves your productivity, which broadly is how many things you can actually get done in a given amount of time. It seems easy enough.

But when I talk to folks who use these apps, I see a strange inconclusiveness. A scant minority of us check off everything every day. An equally tiny minority simply Cannot Even and are curled in a fetal ball awaiting imminent firing. But most of us? We’re just sort of … meh. We bounce from app to app, never quite finding a home. “I’ll try that one. I’ll try that one. I’ll try that one. Maybe this will do the magic!” as Randy Shulman, editor and publisher of Metro Weekly, Washington, DC’s LGBTQ paper, tells me. Sure, we’re getting work done! But we always feel slightly out of control, haunted by the to-dos at work and home that we just aren’t nailing.

The question is, why? Not just why it’s so hard to make a to-do app that works, but why people often feel so distraught by their hunt for the perfect organizational system. I’ve written about software for years, and I can tell you that people often have surprisingly deep feelings about their apps. But rarely is a category of software linked to such vistas of despair.

‘Vistas of despair’ is a bit much, Clive. Consider that the purpose of a task is not to be checked off: it’s a bit of information about us and our work. That might turn down the angst.


We are seeing defections away from baseline task/work management toward communication-centric business operating systems, content-centric work processing tools, highly flexible spreadbases, and project management solutions that can scale to support massive efforts of thousands of projects interconnected in networks with hundreds of thousands of tasks. These alternatives support very different constituencies who are trying to accomplish very different things. In general, core task/work management tools are a bad fit for these constituencies, hence defection.

Thompson digs into one of the common complaints about task management: it's easy to add tasks, but harder to check them off.

One of the most famous productivity systems—David Allen’s Getting Things Done—is ruthlessly focused on rigorous planning and editing of tasks. It can take hours, but once you’ve done that hard work, you can plow through the tasks, one after another, with the metronomicity of a Chrysler line robot.

The problem is that we too often don’t really plan. Digital apps make it easy to add more tasks to the pile, and it feels good to get tasks out of our Zeigarnicized heads. So we do, frenetically.

“We call it snowballing,” says Amir Salihefendić, who founded the app Todoist in 2007; it currently has 30 million users. “They keep postponing stuff. And then suddenly you have a hundred tasks that you need to do.” Weeks or months later, your Todoist app is a teetering ziggurat of tasks, too painful even to behold. Omer Perchik, the creator of another app—Any.do—calls this problem “the List of Shame.”

And then what do we do? You’ve probably done it: We panic, give up, and quit. We “declare to-do bankruptcy.” We toss the list away in defeat and start fresh.

But what if you are working with a group on a project? You can't just declare bankruptcy. You have to remove tasks that no longer make sense, and reorganize other tasks to match what people are actually working on. This is true in individual task management, too.

Give a gift subscription

Thompson also wants to ennoble task management as some spiritual path to making ourselves better:

To-do lists are, in the American imagination, a curiously moral type of software. Nobody opens Google Docs or PowerPoint thinking “This will make me a better person.” But with to-do apps, that ambition is front and center. “Everyone thinks that, with this system, I’m going to be like the best parent, the best child, the best worker, the most organized, punctual friend,” says Monique Mongeon, a product manager at the book-sales-tracking firm BookNet and a self-admitted serial organizational-app devotee. “When you start using something to organize your life, it’s because you’re hoping to improve it in some way. You’re trying to solve something.”

With to-do apps, we are attempting nothing less than to craft a superior version of ourselves. Perhaps it shouldn’t be a surprise that when we fail, the moods run so black.

I believe that managing tasks should be incorporated into the methods and tools we use to think, write, and remember. That might make me a better version of myself: not spiritually, but a me with a better memory.

I am certain that Thompson takes notes of the meetings, calls, and research he is conducting. Perhaps in a spiral notebook. Perhaps notes in a Word document. I believe that the tasks that are part of our 'workings' -- our journals, notes, and other written materials -- should be embedded there. Not in a separate to-do, stand-alone, unintegrated app.

I never have to worry about bankruptcy because I manage four kinds of markdown journal files in my markdown-based 'work processing' approach. And I lay down new journal entries every day, and never delete anything.

Share work futures


This is a short introduction to my Taskora method for ‘work processing’. You can skip ahead if it’s too wonky for you.

Daily Journal -- I create a daily journal file where I capture snippets of things I've read, ideas that have occurred to me, and links to materials I might read in the future. Generally, these items are tagged ('#platform-capitalism', or '#clive-thompson'), and often associated with tasks. When I have finished this analysis of Thompson's essay, and posted it, there will be a task like this in a daily journal file --

- [x] work futures | clive thompson and the dead-end of task management 2021-08-03

-- representing that I completed that task. However, its previous state might have been this --

- [?] work futures | clive thompson and the dead-end of task management

-- with the question mark denoting that the task was provisional, something I might do someday.

At any rate, I seldom worry about how many tasks exist in my journal, nor do I feel the weight of the tasks there. They are milestones, landmarks, and aspirations, not pebbles in my shoes.

Meetings/Calls -- I keep notes during calls, and often create tasks for action items. These I mark with some indicator, like the person's name or the company's name --

- [!] AdjectiveNoun | send an outline of the proposed topic for the November conference

-- and, as in all cases, I rely on the search capability in my markdown platform, Typora, to search for all tasks related to 'work futures', 'AdjectiveNoun', or 'Miller's Plumbing'. This task is urgent, denoted by the exclamation mark ( ‘- [!]’).

Projects -- Once an activity becomes large enough to warrant it, I create a journal file for the project describing what it’s about, pending and completed action items, and so on. Since these change frequently, I often timestamp sections of the project file (newest at the top) and move tasks forward, or indicate their status. This often entails determining that a planned task is blocked, waiting for something else to happen, like this --

- [/] AdjectiveNoun | second panel session is being reconsidered

-- or a task can be made inactive, like this --

- [:] movies | Triumph of Love | although I love Mira Sorvino, I couldn't watch it: it's about a princess.

Source Materials -- I keep large excerpts or complete copies of material I plan to refer to and reread. Can include tasks, but always includes a lot of tags. (This is part of my streamlined zettelkasten method, too detailed to explain here.)

One thing to note: tasks in the Taskora convention are spread all over the place. I don’t have long lists of tasks all by themselves. Tasks are embedded in the text that I use to explain, plan, consider, and remember. I create nothing like a ‘list of shame’. Tasks are just a special form of notes that I leave behind in my daily meanderings, and that I can find again by search.

When I want to see all the tasks associated with AdjectiveNoun, I search for

] AdjectiveNoun |

and Typora shows the tasks in the search results, no matter where they exist in my markdown files. I don’t care where they are, because I can always find them. So, there is no ‘list of shame’ staring back at me.

If I would like to see all provisional tasks for all projects, I would search for

- [?]

which would show a long list, but Typora’s search results show the resulting lines in the files so I could click on ‘- [?] work futures | clive thompson and the dead-end of task management’ to get to a specific task in a specific document.

[Note that the way I denote tasks is not ‘supported’ in Typora: it’s just a convention using text. I am actually avoiding the binary tasks supported directly in Typora (covered in more detail in the Taskora Convention).]

I would never declare task bankruptcy because tasks are woven into my workings, the way that I think, write, and remember. Bankruptcy is just not a consideration.

Yes, projects come and go. Yes, I create aspirational tasks ( ‘- [?]’ ) that I may never return to. So what?

At the end of a year (usually a few months into the new year), I create a new folder for the entries for the new year (like 2022, say next April) and move all the markdown files — journal entries, meetings/calls, and still-active projects — into the 'Journal 2022' folder. The 2021 items stay in their 2021 folder. This minimizes the search scope when I am looking for active 'work futures' or 'AdjectiveNoun' tasks. I can always open the folder of an earlier year, or all previous years, if I need to.

A long digression, but intended to show that content-centric task management — and the Taskora Method — sidesteps some of the ills of other approaches.


Thompson describes various approaches he's used to manage tasks and they share one glaring characteristic: they were isolated from the tools and methods he uses to keep track of everything other than tasks.

For years, I had a very rudimentary to-do system. Using a piece of paper, or maybe a document on my PC, I’d list my main areas of work (“WIRED Column,” “Household,” and so on). Then I’d write out all my tasks under each heading. (Under “WIRED Column”: “Call scientist about study.”) Finally, I’d make a plan. I’d number all my subtasks. Typically I’d hopscotch from project to project: My number one task would be the fourth item under “Household,” then number two was the seventh item under “WIRED Column,” and so on. Finally, with my plan laid out, I could power through my list.

Or at least I’d try to. Sometimes my system would work for days or weeks, but eventually it’d balloon into a List of Shame, and I’d guiltily declare bankruptcy.

Then he wrote an app.

One evening a year ago, I sat down and bashed out a prototype. The next day I started using it and found, to my delight, that it worked much as I’d hoped. I now had a numbered list I could sort and unsort quickly. I used it every day for months. Projects came and went; I filed stories and juggled tons of household errands. It felt lovely to have a tool designed for precisely the way my mind worked.

The thing is, it didn’t improve my productivity. It certainly did not increase how much paid work I accomplished. I was still filing the same number of stories, and doing the same life chores, in the same amount of time. I still found myself getting piled up and spiraling into to-do bankruptcy.

Sure, I could visualize my tasks better. But that didn’t move the needle on my efficiency. In fact, one day while working on the very story you’re reading now, I found myself staring at a monstrous List of Shame in my app. I declared bankruptcy, and then I shakily pulled out a single piece of paper and reprioritized, writing down a small handful of things I could actually accomplish.

I still use my app, intermittently. But building it made me realize a grim fact about to-do software, which is that even the most bespoke, personalized version couldn’t unfrazzle my mind.

I am not trying to increase my productivity, in any direct way. I am simply trying to keep track of everything I am up to and tasks are just one aspect of that.

Thompson never describes how he does everything else. Is it a bound journal and a nice fountain pen? A tape recorder? Word docs?

I think his problem is that he is cleaving his workings in the wrong way. But instead of attacking that, and bringing it all together, he philosophizes about the nature of time, or how we are sending messages to our future selves through tasks.

Ryder Carroll, the creator of the Bullet Journal paper-based method for organizing your work, puts it in even more starkly existential terms. “Each task is an experience waiting to be born,” he tells me. “When you look at your task list that way, it’s like, this will become your future.” (Or if you want the European literary-philosophical take, here’s Umberto Eco: “We like lists because we don’t want to die.”)


He finally considers an alternative approach or two. One is 'time blocking' using calendars:

Instead of putting tasks on a list, you do “time blocking,” putting every task in your calendar as a chunk of work. That way you can immediately see when you’re biting off more than you can chew. Cal Newport, a computer scientist at Georgetown University and guru of what he calls “deep work,” is probably the staunchest advocate of time blocking. “I think it is pretty undeniable that time blocking, done well, is going to blow the list method out of the water,” Newport tells me. He says it makes you twice as productive as those suckers who rely on lists. Time blocking forces us to wrestle directly with the angel of death. It’s natural that we then screw around less.

Well, ok. Except I don't want to wrestle with the angel of death. I just want to keep track of things and to find what I need when I need it, which is dozens of times a day.

Thompson thinks that maybe good-old-paper is the answer, but he's asking the wrong questions:

A whole bench of task-management philosophers believe that the best interface isn’t digital at all—it’s paper.

Paper forces you to repetitively rewrite tasks, as when, say, you transfer all last week’s undone to-dos to this week’s list, or when you erase and rewrite calendar events. That’s what I do when the productivity software I wrote for myself fails me. “Making that choice over and over again,” Carroll tells me, “is the first opportunity where you’re like, ‘Why am I doing this?’” The inconvenience can be clarifying. Making a list on a sheet of paper is an unusually rich metaphor for life: It takes effort, and the space fills up more quickly than you expect.

The usefulness of paper here cuts to the real heart of what makes to-do management such a grim problem. Apps, lists, and calendars can help us put our priorities in order, sure. But only we can figure out what those goals are. And setting limits on what we hope to do is philosophically painful. Every to-do list is a midlife crisis of unfulfilled promise. Winnowing away things you’ll never do in a weekly review is crucial, yet we dread it for what it says about the boundaries of existence. Our fragile psyches find it easier to build up a list of shame, freak out, and flee.

This is what makes to-do software unique. The majority of tools we use in our jobs are about communicating with someone else. All that messaging, all those Google docs, all that email—it’s about talking to other people, documenting things for them, trying to persuade them. But a to-do list is, ultimately, nothing more or less than an attempt to persuade yourself.

Why waste time winnowing away tasks? I can understand the desire to review the status of the AdjectiveNoun project, but why throw tasks away? Just change their state, if necessary. Make a task urgent, make another inactive, create a new provisional task. But why delete any? It’s all information in the stream of work.

Thompson never asks himself 'how do I keep track of everything else I am thinking, doing, and want to remember? Can I simply insert tasks into that stream of activity?'

I think the answer to that question is ‘Yes’, but Thompson never even discussed how he does everything else. Although he obviously does it somehow. Why not smoosh it all together and stop worrying about productivity?

Me, I don't care about increasing output, I just want better outcomes. And a better memory.