Hundreds of books had been added to a WeChat Read bookshelf, but never properly sorted into groups. In one organizing session, the book list was brought to a local device, an Agent proposed categories, and the user confirmed them. Then, using Computer Use, the Agent operated the app on a Mac and moved the books one by one into their corresponding groups. According to the record of this session, the bookshelf was organized in a little over an hour.
The task soon prompted a debate: How would the books be used once sorted? If it did not lead to reading even one more book, was it false productivity? Another interpretation was much simpler: this was just tidying up.
These two judgments may seem to conflict, but they measure different outcomes. One requires organizing to lead to later output; the other treats restoring order as the task itself.
First, separate what that hour accomplished
What this task clearly completed was categorizing and putting the books away. It did not provide book-finding times before and after, record subsequent reading, or establish how many hours of manual work it saved. Calling a little over an hour of execution “a little over an hour saved” would change what is being compared: if the user would not have organized the books by hand in the first place, there was no equivalent amount of manual work being replaced.
But the fact that someone would not have done something before does not mean it is not worth doing. Many personal tasks are put off for a long time, not because they are meaningless, but because their value is not enough to outweigh the effort required at the time. When tools lower that cost, reorganizing a bookshelf or turning saved items into a list may become worth trying.
The value added here is an outcome that had not been achieved before. It may be useful, or it may simply be satisfying. Either way, it should be weighed against the cost, but there is no need to force it into a measure of work efficiency.
Categories should support the next use
A bookshelf can be organized by discipline, author, or reading progress, or by the question someone is currently researching. None of these approaches is better apart from its purpose. Someone preparing to write an article may need a very different way in than someone looking for a weekend read.
Research on personal information management has long focused on how things that have been saved can be brought back into everyday activities. In introducing the Keeping Found Things Found project, William Jones connected effective information management with everyday planning. This perspective offers a useful reminder: a system of categories can be designed backward from what someone plans to do in the future, rather than simply aiming for a tidy-looking directory.
Suppose someone is researching urban life. A history of architecture and a novel might belong in the same group; separating them strictly by publication category could instead add steps to finding them. Conversely, if the purpose is simply to browse the whole collection, a stable, simple classification by discipline may be enough.
An Agent can lower the cost of proposing a categorization scheme, but the user still needs to recognize the categories. A system that a machine can explain in full, but that the user cannot remember the next time around, is unlikely to serve as a lasting point of entry.
Tidying has an endpoint; organizing can also become procrastination
Recognizing the pleasure of tidying as a value does not mean every organizing task is worth continuing. Once the categories are in place, someone may want to standardize the covers, fill in missing descriptions, add multiple layers of tags, and then make a reading plan. A task that started out with a clear limit can keep expanding.
If the goal is to get the messy bookshelf into a state that feels good to look at, the task can end when it reaches that state. If the goal is to make reading more convenient, it is worth returning to actual use: Can someone find a book when they need it? Can a newly purchased book be put away easily? Do the categories still need frequent reworking?
These questions can be answered through later use; there is no need to build a complicated dashboard to prove productivity. Observing a few real book-finding situations is closer to the goal than counting how many new tags were added.
If an organizing session brings only a sense of satisfaction, it can simply be called organizing. The real concern is when ever-expanding preparation takes time away from what someone originally wanted to do, while efficiency remains the explanation for it.
Personal agents do not have to do only things that fit in a weekly report
Commercial tools often explain their value in terms of hours saved and revenue gained. Personal life also includes interests, comfort, a sense of order, and exploration, which may not correspond to output metrics. Spending time making a personal book list, like spending time arranging physical books, can be guided by the user’s own preferences.
Still, that value does not exempt a product from evaluation. Whether one enjoyable organizing session is enough to justify a long-term subscription is a separate question; nor does a categorization looking good at the moment it is completed prove it will remain useful. One-time satisfaction and continued use should be observed separately.
For this bookshelf-organizing session, the most accurate conclusion remains specific: the user completed an organizing task they wanted to do, and any benefits to subsequent reading have not yet been recorded. The value of a personal Agent can begin here, without first packaging every kind of personal wish as productivity.
Related materials
- Usage record: WeChat Read bookshelf categorization, confirmation, and interface operations
- Time spent on this organizing session and discussion of tidying as a purpose
- William Jones: Putting our digital information in its place, introduction to the Keeping Found Things Found project, consulted on October 19, 2026.