AI, data, cybersecurity or mobile? Choose by the problem

A popular field is still the wrong starting point if you dislike its daily work

The conclusion

• Translate one Kenyan work problem into four different technical tasks.
• Confirm the foundations and device requirements before committing.
• Use one bounded project to notice which difficulty you want to pursue.

1. Separate the work hidden behind each label

Take a service you understand, such as a retailer handling customer requests. Data analysis asks what patterns the records reveal. Mobile work asks which task belongs on a phone and why. Cybersecurity asks how identities and customer information must be protected. AI is relevant only when prediction or generation improves a defined decision and its errors can be managed. The same setting exposes different daily responsibilities.

Use one Kenyan retail scenario to separate the fields. Data analysis can explain sales patterns; mobile work can make an essential task accessible on a phone; cybersecurity can protect accounts and customer information; AI is relevant only if prediction or generation is justified by suitable data and tolerable error. The problem reveals the role more clearly than a list of trends.

Add a “not this field when…” line beneath each option. AI is not the first answer when a transparent rule solves the task. A mobile application is not automatically useful when users lack a reason to install it. Data analysis cannot repair unreliable collection by itself, and cybersecurity is not a one-time checklist. These boundaries make the comparison more realistic than assigning every problem to a fashionable label.

2. Check what the first month really assumes

Read the provider’s prerequisites, assessments and software requirements. A mobile path may expect programming and interface basics; security may assume networking; data and AI may need statistics or Python. Also confirm that your device can perform the labs. Use a short local or free resource to test a missing foundation before buying depth. Discovering the starting floor early is preparation, not evidence that you do not belong.

Inspect the real starting floor. Mobile development may require programming and interface basics. Cybersecurity may require networking and operating-system concepts. Data and AI routes may assume spreadsheets, statistics or Python. Spend a short week testing the assumed foundation. The pause prevents an advanced syllabus from feeling like evidence that you do not belong.

A prerequisite week should end with a small artefact: a cleaned table, a network sketch, a simple mobile screen or an error analysis. Save it with a note about what required help. You are testing more than comprehension; you are testing whether the work fits your available device and connection. That evidence can prevent an expensive route from becoming the place where a basic constraint is discovered.

3. Run one fair test and listen to your response

Keep the tests comparable in time and scope: clean an open Kenyan dataset, sketch a secure account flow, build a simple mobile screen, or evaluate a rule-based baseline before an AI model. Stay long enough to meet an error. Notice whether you want to investigate, explain, protect or redesign. Save the artefact and reflection. That felt response gives better career information than another list of fast-growing roles.

Build the smallest comparison project and keep the constraints equal. Notice which difficulty draws you back: tracing a security weakness, cleaning a messy table, fixing a mobile interaction or evaluating a model’s error. Your answer can change later. For now, it gives the next 90 days a direction rooted in work you have felt, not a claim about the hottest occupation.

Ask a practitioner or informed peer to question the test without selecting your career for you. They can identify a missing foundation or an unrealistic task, but only you can notice which work you wanted to continue after the conversation. The direction you choose should feel specific enough to plan and flexible enough to revise after the next project, because no 90-day experiment needs to become a lifelong label.

Summary

A familiar problem separates the roles, prerequisites reveal the true starting floor, and a fair project shows which difficult work deserves another 90 days.

To move closer to the person you want to become

• Write one work problem in four field-specific versions.
• Check the software, device and knowledge required by each route.
• Complete one bounded test with lawful data.
• Record which error you wanted to understand more deeply.