Pipeline tells you exactly and only what you need to do next.


I’ve been working on systems to keep track of what I’m doing and what I’ve done since the time it was important to others that I do anything (before that I just blundered around doing whatever came to mind).  Over the years I’ve developed a number of manual, automatic, physical and electronic systems, each one a refinement of the previous.  During this time I’ve learned a lot of about what works and what doesn’t (at least for me) and based on my experience working with other people and systems, I think that I’m not alone in this quest for such a system.

The latest incarnation of this is something I call “Pipeline”, and it’s currently in the design phase.  Simply put, at any given time Pipeline presents you with the thing you need to work on next, allowing you to focus on only two things at a time:

  1. What you are doing now

  2. What you are doing next

There’s two specific advantages to this myopic approach.  The first is that there is zero ambiguity about what to do at any given time; as soon as you finish what you’re doing, you consult Pipeline and it tells you exactly what to do next.

The second is that there is almost no cognitive overhead associated with thinking about anything other than the task at hand.  Worst-case you’re distracted by thinking about the next task (if this becomes a problem Pipeline could even hide that task until you’ve committed completion of the current one, nut my experience leads me to exposing the next task whenever the system is consulted).  What this prevents is meta-thought about the prioritization of all tasks, and a form of premature optimization that can be crippling (and is rarely beneficial).

Where do tasks come from?

The “next” task is drawn from a (if you’re like me, bottomless) pool of tasks that are populated ad-hoc by you or anyone else you trust.  Pipeline’s job is to handle prioritizing these tasks based on any source of information it can draw from.  I’m not ready to go into detail about this part yet, but suffice to say it’s intentionally simple, and leans heavily on a negative feedback loop, like all good cybernetic systems.

What if I can’t do the next task?

Sometimes Pipeline will pick a task you can’t (or don’t want to) do right now. In those cases you simply “decline” the task, and it’s returned to the pool. Declines are recorded, along with any information about the context of the decline that are available and this negative feedback is used to influence the prioritization of all tasks.  Pipeline doesn’t allow of infinite procrastination though, and at some point you’ll be forced to accept a task or it will just keep coming back*.


As mentioned earlier, Pipeline is the latest incarnation of a system I’ve been developing and building for decades, and will continue to evolve over time.  I do have an implementation in mind for this specific iteration of the idea, but not a specific implementation timeline.  That said, if you’re interested in helping out tweet me and we can talk about some of the implementation-level details I have in mind and maybe write a little code.

- Jason