Can AI make your To-Do Lists Smart

To-do lists are great productivity tools that help the busy professional to remain organized. Yours truly is a committed user of ‘Tasks’ - to-do list management software which comes along with the GMail Suite. Before ‘Tasks’ came, I have used Evernote and StickyNotes for managing to-do lists. I have found digital ‘To-Do Lists’ a boon for keeping track of both professional as well personal tasks. No doubt, many others find the same.

However, it turns out, ‘To-Do Lists’ can have a dark side.

As productivity expert and author Nir Eyal explains in this interview, ‘To-Do Lists’ can create a ‘Tyranny of the Small and Insignificant’. Because it feels satisfying to tick-off items from a to-do list, one can gravitate towards doing more of the ‘smaller’ tasks rather than the ‘bigger / more important’ tasks. The ‘satisfying feel’ is actually linked to a very powerful physical mechanism of dopamine release in the brain. I do feel this ‘Tyranny of the Small and Insignificant’ happens at times with all to-do list users and you need to consciously prioritize among the tasks. However, that imposes a degree of additional effort.

While Nir Eyal advocates avoiding ‘to-do lists’ altogether in favor of a different approach, many people might still find them useful and are habituated to using them. The fact that they are linked to a fundamental brain mechanism makes them powerful productivity tools, as many have experienced. One way in which I believe it is possible to ‘fool / leverage’ this ‘feels satisfying’ internal mechanism yet get your most important tasks done without consciously spending effort in prioritizing them is to break them into sub-tasks and put them on your to-do list. That way, as you complete the sub-tasks, you still get to tick them off and get the ‘satisfying feel’ and you get to tick off a lot of items in your list when you do the entire task. I have tried this and see it working well. This article also echoes the same idea. As illustrated below, it can be about listing sub-steps of a complex task or sections of a document as sub-tasks.

However, it would be still better if somehow the importance of the task could be reflected in the to-do list and somehow the ‘satisfying feel’ could be increased once one does that. One can color code tasks and use multiple lists but that can be a bit cumbersome. One mechanism which might work is a to-do list where the size of the task tab is larger for more important tasks - that way, when they are ticked off, the length of the remaining list shortens by a larger degree and the ‘satisfying feel’ is increased. 

It is probably not very difficult for the makers of to-do list software to provide a feature to enable users to do this. However, it can be still more wonderful, if the software could somehow smartly sense the importance of the task as it gets typed and set the size of the tab accordingly. That will be a truly ‘smart to-do list’. The artificial intelligence technique of NLP (Natural Language Processing) can be employed to do this. The problem is a bit like the well-known NLP problem of sentiment analysis where a text input needs to be mapped to a sentiment. Here, the mapping will need to be to an importance level. The models should ideally be personalized to each user and hence the user will need to provide some initial examples of tasks along with their importance levels. Since the user can also adjust the tool suggested importance levels, the models can continuously get refined. Powerful language models like the recently released GPT-3 by OpenAI, among others, could be employed to build this ‘importance sensing’ models.

I believe a lot in the idea of AI acting like superpowers for humans. The above is a speculative idea about how AI can potentially help in enhancing productivity.

Banner Photo by Suzy Hazelwood from Pexels