A large goal is emotionally attractive and operationally vague. “Start a business,” “change career” or “write a book” can create motivation without telling you what to do on Tuesday.
AI can help convert the idea into a 90-day experiment—but only if you make it show its assumptions and keep the important decisions for yourself.
Define the change in observable terms
Begin with an outcome you could verify after 90 days. For example:
- speak to 20 potential customers;
- complete and test a first portfolio;
- write 24,000 words;
- apply for 12 carefully chosen roles;
- run three small versions of a service.
Avoid asking AI to decide whether the goal is “right.” Give it the practical job of organising evidence and options.
Ask for questions before a plan
Use a prompt such as:
I want to test this goal over 90 days. Ask me ten questions about time, money, skills, responsibilities and evidence of demand. Do not create the plan until I answer.
This interrupts the tool’s tendency to produce a generic twelve-week schedule from almost no information.
Separate facts from assumptions
Ask the model to create three lists:
- What is already known.
- What is being assumed.
- What must be tested.
If the goal involves money, employment, health or legal obligations, verify claims through primary or professional sources. AI output is not evidence simply because it is specific.
Design the smallest useful tests
A 90-day plan should create information, not just activity. Early actions might include a customer interview, a sample project, a price test or a conversation with someone already doing the work.
For business goals, AI News & Updates’ guide to creating a business plan with ChatGPT shows how to use the model for structure while independently checking market size, competitors and financial claims.
Build in decision points
Set reviews at days 14, 30, 60 and 90. At each review, record:
- what was completed;
- what evidence changed;
- which estimate was wrong;
- what should stop;
- the next smallest test.
Ask AI to summarise the log, but compare the summary with the original entries. Do not allow a confident narrative to hide disappointing evidence.
Keep a no-AI version
Write the central goal, boundaries and stopping rules in your own words. If the tool disappears or changes, the experiment should still make sense.
AI is useful when it reduces the clerical load of planning, comparison and review. It is harmful when a polished schedule replaces honest thought.
A good 90-day experiment does not promise that the big goal will succeed. It gives you better evidence about whether—and how—to continue.
Editorial disclosure: AI News & Updates is a companion publication. The linked guide is included because its verification approach directly supports this planning method.
Leave a Reply