mtmail
3 months ago
I worked on a todo-list app project (to sell ads, didn't work) and saw how the average user manages their tasks. I'd say "flowers mom", "new tyres", "trash day" or sometimes nothing but a name or phone number. An AI would have to make a lot of guesses. I'm doubtful prioritizing tasks against each other works without a huge amount of context. An AI could still help by asking questions back to the user "Is A more important than B?", "What's the timeline on C?", "Can you break down D into actionable sub-tasks?". The user would spend more time in the tool, not less.
Running4M
3 months ago
Hello mtmail,
Thanks for sharing your thoughts!
I completely agree that for simple tasks like a grocery list or reminders (e.g., "buy flowers" or "trash day"), just a word or two is enough to jog the user's memory. Those kinds of tasks don’t really need much detail or input from an AI.
What I had in mind for integrating ChatGPT, though, was more for work-related or hobby-focused tasks—things that require a bit of planning or creativity. As you mentioned, for ChatGPT to be effective at breaking down or prioritizing a task (or at least offering tips or starting points), the user would need to provide more context or detail.
To make this easier, I imagine the user writing a brief description of the task with the help of prebuilt prompts. These prompts would guide them in crafting the description so that ChatGPT can clearly understand what the task is about. For example, prompts could ask, “What’s the goal of this task?” or “What’s your timeline?” This way, the process of describing tasks wouldn’t feel overwhelming or time-consuming for the user.
I get that this might require more effort than just jotting down "do laundry" or "call mom," but my focus is on helping with more complex tasks—like writing, research, or work deadlines—where this extra input could genuinely help users stay ahead of their schedule and avoid procrastination.
What do you think about that approach?