Which tool to choose
The tool is the conversational assistant, used to break things down: you give it the objective and it splits it into phases and steps. Keep the plan in a document, because a project lives over time and the plan needs updating as you go. For deadlines and reminders, a calendar or a management app completes the picture: AI structures, the tool keeps track of progress. The judgment part — how much time it really takes, what's realistic for you or the team — stays yours: the assistant draws the map, but you know the real distances.
How to do it
- Explain to the AI the final objective, clear and concrete, and the constraints (time, resources, people).
- Ask it to break it into phases and concrete steps, with the right order and the dependencies.
- Calibrate the timing to reality: AI tends to underestimate, so add a margin.
- Put the milestones in a document or a management tool and update it as you progress.
A concrete example
Davide wanted to launch a small website for his business, but the project seemed like a mountain and he never got started. He explained the objective to the AI and asked it to break it into steps. The assistant gave him a sequence: decide the content, choose the tool, write the texts, set up the pages, publish. Suddenly the mountain was a staircase. Davide stretched the timing a bit, which the AI had set too tight, and put the milestones in his calendar. He started from the first step instead of staying stuck in front of the whole: the project had become one thing at a time.
When it does NOT work (and how to fix it)
If the timing is too optimistic
AI almost always underestimates how long it takes. Add a margin to each phase and remember the unexpected, which always happens. A plan with unrealistic timing only creates frustration: better wider but respectable deadlines.
If the steps are too big
If a step is still a mini-project, it doesn't unblock you. Ask the AI to break it down further, into moves you can do in one work session. The smaller and more concrete the first step, the easier it is to start.
If the plan doesn't survive contact with reality
No plan holds up intact: obstacles emerge, priorities change. Treat it as a living document, to update when things change, not as a carved-in-stone table. A plan exists to give you direction, not to cage you in forecasts made at the start.
A tip from someone who really uses it
Have the AI give you the smallest possible first step, and start from there. Most projects don't die from lack of a good plan, but because the whole is frightening and you never begin. The value of breaking things down isn't having the complete map, it's having a first move so small and clear that it can't be put off. When the assistant returns the sequence, don't look at the whole staircase: look at the first step and do that. The project advances one step at a time, and the hard thing — starting — becomes trivial when the first step is "write the list of content" instead of "make the website". AI is excellent at making small what seemed enormous.
Frequently asked questions
Can AI make the plan on its own?
It can build an excellent breakdown, but the timing and the real feasibility you calibrate yourself: it doesn't know your constraints and tends toward optimism. Use it for the structure, and adjust to reality.
How do I manage the dependencies between steps?
By asking the AI to indicate what must come before what. A good breakdown shows the order and the steps that block the others, so you know where to start and what you can't skip.
Should I update the plan?
Yes: a project lives over time and the plan must be kept alive. When obstacles emerge or priorities change, update it. A rigid plan that no longer reflects reality stops being useful.
If I have the AI's plan, is the project taken care of?
No, and that's the difference between planning and doing. A plan, however well made, is only a map: the project advances because you execute the steps, manage the unexpected and keep the plan updated along the way. AI removes the initial paralysis and gives you a clear sequence, which is a lot, but it doesn't walk for you. The risk is mistaking a nice plan for a started project and staying still perfecting it instead of beginning. Have it give you the breakdown, then focus on the first step: it's the action, not the map, that carries a project to the finish line.