Greetings, data-tamers.
How many times have you faced an enormous pile of tasks, appointments, and responsibilities--whether in your digital life or otherwise--and found yourself unable to decide what to work on? Everything feels important, or urgent, or both; nothing has enough "give" to postpone or cancel. There's always a reasonable justification, a legitimate need to do each thing. Obviously, otherwise it wouldn't be on your list in the first place!
This past week, I accidentally stumbled into an interesting way to attack this problem. I won't call it a "solution," a label far too generous to apply just yet, but I'm encouraged by how it's working for me so far.
I've been experimenting with connecting my to-do list and calendar with an autonomous AI agent, first to perform audits of how I'm actually spending time, and subsequently to make recommendations for how I should spend time more efficiently. I'm doing this for two reasons: one, I've never set up an AI agent like this before, and this is a great hands-on way to learn how to do it and how it works. And two, if I can get it do what I want, it will add immediate tangible value to my life.
Now, there's no magic bullet here; you can't just throw AI at a problem willy-nilly and get a good result. These tools require specific, targeted context in order to produce a desired outcome. So, I approached the problem as through I were hiring a very smart personal assistant to take over time management for me. What would I have to tell this person to ensure he could intelligently steer my activities?
...and that question is the unexpectedly valuable approach.
Because as hard as it is to prioritize our own tasks sometimes, we're forced to think about the problem in a very different way if we have to delegate that responsibility to someone else. What would they need to know in order to make good decisions?
- How long will the tasks take? (and how wrong are my time estimates, historically?)
- Are there any actual deadlines?
- Which tasks require uninterrupted focus, and when am I likely to get that?
- Which tasks make sense to bundle together as a batch?
- Which tasks require physical effort?
- Which tasks can only be done in a specific location?
- Which tasks depend on other people?
- Which tasks depend on other tasks being done first?
- Which tasks should go on the back burner for some unknown future date?
- How often should I review the "master" task list to ensure nothing falls through the cracks?
- What happens if some of these don't get done at all?
- Do I enjoy any of these tasks enough to consider them "fun" or otherwise rewarding?
- Are there any specific tasks or categories I really dread?
- How do my energy levels fluctuate throughout the day?
- Am I allowed to be interrupted to deal with "fires" that pop up?
If your assistant knew the answers to all of those questions, they would have no trouble telling you what to work on.
Of course, these questions are not always easy to answer. But I'd bet you don't often think about them in such specific terms, and you never think of them in the context of delegating the job to someone else.
Just walking through the mental exercise is often enough to shake some new ideas loose; you don't need to hire a personal assistant. Heck, you don't even need to mess with an AI agent like I'm doing. Just imagine you're going to put someone else in charge of your calendar and to-do list: your spouse, your friend, your kid, your cat. Anyone. What would they need to know in order to help you accomplish your goal effectively? Talk through it out loud if it helps. (Especially if it's your cat.)
My experiment here is with time management, but this also works just as well for tasks like cleaning your email or organizing your photos or documents. You just have different questions to answer to provide the right context. For example, if you wanted someone else to help you get to Inbox Zero, what would they need to know?
- Which emails require a response?
- Which emails require follow-up action, even if not a response?
- What should I do with emails that have information you need to know, even without a follow-up action?
-
What should I do with emails that have information you want to know, for fun or curiosity?
- What should I do with emails that you want to be able to find later if needed, like order confirmations?
- What should I do with emails that are only useful for short timeframes, like shipping confirmations?
- Which emails are simply trying to sell you something?
- Which emails never get read even with the best of intentions? (Spoiler: unsubscribe!)
You can do the same thing with photos and a different set of questions. Again, the trick is to imagine what you would have to tell someone else in order to have them complete the task for you. It won't eliminate the challenge of deciding, but it brings the important questions to the surface very quickly.
Have you ever tried something like this? Did it work? Or if not, does it seem like it might help your situation? Let me know! I'm going to keep playing with this AI agent to see if I can coax real value from it.
Until next time, happy data-taming!