Are priorities stupid?
Dates already capture importance. I'd rather use AI to review and replan my work than maintain priority labels.
Jamie Davenport
Priorities don’t belong in a personal productivity system that supports dates. The date is what captures the importance. Something scheduled for today is higher priority than something without a date.
High, medium, low. P1, P2, P3. Each asks me to make another decision about a task, then I have to keep that decision up to date. I’d rather decide when to do the work and get on with it.
A personal task list spanning work, business and life, with priorities, dates and effort estimates. Reviewing pricing is high priority but undated: what does this mean? Right now or is tomorrow okay? Finishing a SOC 2 security review is medium priority and due today: should I do this first? Rebuilding a revenue chart is low priority and takes a week: why am I even bothering with this? The list also includes sending a client proposal tomorrow, renewing car insurance on Friday and reorganising old photos without a date. Thoughts appear in turn while the list stays still.
Background
I use my own productivity tool, 
Sidequest, the culmination of many decisions to make getting things done easier (for me). One of those decisions: no feature gets added until there’s a burning need for it. After prolonged use, I’ve not missed priorities. So why are they such a standard feature?
Todoist has four priority levels, and TickTick includes an Eisenhower matrix. The intention is to help you identify what deserves attention. A personal task list doesn’t need that extra classification.
I’m both the person choosing the work and the person doing it. A product manager recording priorities to communicate an order of work to a team is solving a different problem.
Dates already capture importance
A task becomes more important as its date approaches. I enter the date once; time progresses, and I don’t need to change anything. A priority label stays exactly as I left it until I remember to change it.
High priority doesn’t tell me whether something needs doing now or next week. Medium priority doesn’t explain whether it comes before a task scheduled for today. The labels leave me with the same decisions about when to act, plus another field to maintain.
Leaving everything high priority from the beginning is hardly better. A task can sit there for weeks, apparently demanding attention. I either learn to ignore the label or feel bad about it. My productivity system shouldn’t give me another thing to feel behind on.
David Allen makes a related argument in Why ABC Priority Codes Don’t Work: circumstances change too quickly for fixed rankings to remain useful. He still considers priority when choosing what to do.
Even Todoist’s priority descriptions point towards scheduling: act immediately, schedule it, or leave it for later. Why stop halfway through that decision? Pick a day.
Dates need maintenance too. Plans change, and an overdue list can become just as unpleasant as a list of permanently urgent tasks. But moving a date makes a new plan. Changing P2 to P1 still leaves me deciding when to do the work. Changing a date captures more information than changing a priority.
Not everything needs a date
I’m happy to keep some things undated until I want to give them attention. That is a deliberate decision which implies that the task isn’t very important. When it becomes important, it gets a date.
The Eisenhower matrix separates urgency from importance and recommends scheduling important work that isn’t urgent. I agree with that action. Similarly, Leo Babauta’s Most Important Task approach deliberately makes room each day for a personal goal. Choosing a day to work on something already gives it that attention.
Skip low priority work
If something takes substantial effort and matters very little to me, why keep it? A low-priority label can let me avoid admitting that I’m never going to do it.
Delete it. Elimination is one of the Eisenhower matrix’s most useful recommendations. My task list doesn’t need to preserve every intention I’ve ever had.
Enjoyment and learning count as value. I’m happy to spend time on a hobby or a long-term project. I’m questioning the obligations I don’t enjoy, don’t need, and don’t expect much benefit from.
Following the same principles gives me a simpler task list, without maintaining the matrix.
The same personal task list, with four more tasks, moves out of the Eisenhower matrix. The SOC 2 security review keeps its date of today, the client proposal tomorrow and car insurance Friday. Reviewing pricing gets a date next week. Prototyping offline mode and planning a walking holiday initially remain undated. Booking a meeting room goes to a teammate via Slack; booking team travel goes to the office manager via email. Rebuilding a revenue chart and reorganising old photos are dropped. A week later, the first three tasks are complete. The pricing review and offline prototype are now planned for today, and holiday planning for tomorrow. The animation loops while visible.
Urgent
Not urgent
Important
Less important
Finish SOC 2 security review
Send client proposal
Renew car insurance
Review pricing
Prototype offline mode
Plan walking holiday
Book meeting room
Book team travel
Rebuild revenue chart
Reorganise old photos
Slack
Bin
Pick something and start
When several tasks have the same date, I scan the list and start something. The choice depends on where I am, how much time I have, and what I feel capable of doing. I don’t need to record that judgement before acting on it.
Dates also narrow the list. I’m looking at the work planned for today, rather than comparing every task I’ve ever saved. If today’s list is too large to be useful, I’ve put too much on it and need to replan.
At the end of the day, I move unfinished work to suitable future dates and spread it out. I can do that manually or ask an assistant. I’m making a new plan for the work that remains.
AI does the heavy lifting
An AI assistant connected to my task list can read tasks and update their dates. It can already translate between implicit priorities and dates.
An illustrative chat continuing from the previous illustration, a week later. Asked what's top priority, an assistant fetches the three remaining tasks: a pricing review planned for today, an offline mode prototype also planned for today, and planning a walking holiday tomorrow. The user says the pricing review isn't important anymore. The assistant asks whether to reschedule it or delete it. The user chooses next week. The assistant updates two tasks: it moves the pricing review from today to next week and brings holiday planning forward from tomorrow to today. The offline prototype stays today. It translates a change in priorities into dates without a priority field.
I can talk about priorities without maintaining a priority field.
Every extra field has to justify the attention it asks of me. Priorities still haven’t earned their place.