What gets replaced
The maintenance layer. Projects, tags, priority flags, filtered views and the weekly review exist to help you find what matters. If something else determines what matters, that scaffolding stops earning its keep.
This is the part people actually abandon apps over — not the storing of tasks, but the upkeep of the structure around them.
What doesn't get replaced
Capture. You still have to get things out of your head, and no model can read what you haven't written.
Storage. Tasks still need to persist between sessions.
Judgement. Deciding what you're willing to let slip is not a task you can delegate to software.
The honest limitation
An AI ranking is only as good as the dump it reads. Write four vague lines and you get a ranking of four vague lines.
It also has no memory of what you didn't say. The colleague who'll be difficult about a delay, the reason a task is really blocked — invisible unless written down.
What a sensible setup looks like
One place to capture, ranking handled automatically, and you doing the work and overriding the order where you know better.
No projects to design, no tags to keep consistent, no review to skip. The list still exists — you just stop gardening it.
Frequently asked questions
Will AI make task apps obsolete?
Unlikely. It changes which parts need a human — capture and judgement stay human, sorting largely doesn't. Expect apps to keep the list and drop the configuration.
Do I still need to write everything down?
More than ever. The ranking quality depends entirely on the input, so a thorough dump matters more in an AI workflow than in a manual one.
Is this just autocomplete for productivity?
The ranking is genuine language understanding of context, deadlines and consequences rather than pattern-completion on a phrase. But it's a convenience layer over a well-established practice, not a new mechanism.