23/08/2026
You cannot build for the future without understanding the context.
This has been particularly clear through my work connected to the UK Space Agency Explore Programme and while developing Aggregator.
We are exploring how Space data, and AI can help organisations.
But the same principle applies whether you are developing space technology, improving a customer experience, planning your career or using ChatGPT on a Monday morning:
AI cannot understand what you have not properly explained, documented or connected.
And more data does not automatically mean better intelligence.
Poor-quality, outdated or incomplete information can produce an answer that sounds impressive while pointing you in completely the wrong direction.
Here are five things I would do before using AI to help solve a problem this week:
1. Define the decision—not merely the task.
Instead of asking, “Can you analyse this?”, explain what you are trying to understand, decide or change.
[Do this now: Write one sentence beginning, “The decision I need to make is…”]
2. Add the relevant context.
Explain the audience, environment, limitations, history and intended outcome. AI does not automatically know which details matter to you.
[Do this now: Add who this affects, what has already happened and what a good outcome would look like.]
3. Check what or who may be missing.
The available information may exclude people, experiences or conditions that were never recorded. This is particularly important for accessibility, invisible disabilities, culture, gender and other forms of representation.
[Do this now: Ask, “Whose experience is not represented in this information?”]
4. Check the source and date.
AI may rely on outdated information, mix sources together or confidently fill gaps with assumptions.
[Do this now: Ask for the source, publication date and any areas that could not be verified. Then check the important claims yourself.]
5. Decide where a human must remain involved.
AI can help organise, compare, summarise and identify patterns. It should not quietly become the accountable decision-maker.
[Do this now: Identify which output can be used immediately, which needs checking and which requires expert or human approval.]
This is systems thinking.
We do not start with the tool. We start with the problem, the people, the context, the data and the decision that needs to be made.
Over the coming week, I’ll be sharing some of the practical ways I use AI across business, innovation, service design, career development and everyday work—as well as how I recognise when it has misunderstood the instruction or produced something that should not be trusted.
For practical guidance on emerging technology, AI and how to apply it across your career, business and everyday life, join my newsletter:
https://kyc-digital-media.kit.com/8a7f48d5ca
You can also follow me on LinkedIn:
https://www.linkedin.com/in/katrina-y-108294b3
