Over the past year, mentoring students through KamiLimu has put me in a lot of conversations with young people trying to find their place in technology.
KamiLimu works with university students in Kenya, complementing what they learn in the classroom with mentorship, practical skills and exposure to people already working in the industry. I have had the privilege of helping lead its Software Engineering Foundations cohort and being a professional mentor, which has meant teaching technical concepts, reviewing projects and helping students think through what comes after university.
But some of the most interesting conversations have had very little to do with code.
Some students are deciding whether they should learn Python or JavaScript.
Some are wondering whether AI engineering is really different from software engineering.
Some want to start businesses. Others want to become really good engineers. Some are simply trying to understand whether the path they are already on is enough.
The questions are different.
But underneath them, I keep hearing the same thing.
Am I moving in the right direction?
I understand that question.
I think most people do, even after they have worked for a few years, changed industries, started a company or built things they are proud of. The details change, but uncertainty has a way of finding new language.
I wrote about something similar in Reinvention Is a Practice. One recurring lesson in my own career has been that progress often requires becoming a beginner again. Technology changes. Industries change. The work changes. Eventually, something you were once confident about becomes unfamiliar again.
Mentoring has made me see the other side of that experience.
It has reminded me that experience is not only about eventually having answers.
It is also about remembering what it felt like not to have them.
Most People Do Not Need More Information
A lot of students already know more than they think.
They have watched the tutorials. They have heard of the frameworks. They know the names of the languages, the companies, the tools and the trends.
What they usually need is not another list.
They need clarity.
They need someone to tell them that they do not have to learn everything at once.
That they do not need to become an expert in every new thing that appears on the internet.
That fundamentals travel.
A programming language is a tool. The principles underneath it are what stay with you.
Once you understand how to think about data, logic, structure, systems and tradeoffs, you are not starting from zero every time you encounter something new.
You are translating.
I have been thinking about this even more in the age of AI.
In AI Is Not Replacing You. It’s Reshaping How You Think, I wrote about how easier code generation shifts the bottleneck elsewhere. Architecture, judgment, problem definition, tradeoffs and understanding how systems fit together become more important, not less.
That is one reason I keep coming back to fundamentals when talking to younger engineers.
The framework you learn today may be replaced.
The ability to understand a problem probably will not be.
That is a much calmer way to learn.
The Best Project Is One You Actually Care About
One lesson I keep repeating is that people learn faster when the work means something to them.
You can build another to-do app. There is nothing wrong with that.
But if you care about photography, build something around photography.
If you care about football, build something for football.
If you care about your community, find a problem there.
One thing I have enjoyed about mentoring at KamiLimu is encouraging students to look beyond technology itself when choosing what to build. Some of our conversations start with code but eventually become conversations about accessibility, culture, informal work, trust, fairness or problems they have encountered in their own communities.
Technology becomes much more interesting when it has somewhere real to go.
The point is not to make every student project revolutionary. Most do not need to be.
The point is to stay interested long enough to encounter the difficult parts.
The broken API call.
The confusing database model.
The deployment that works locally and fails everywhere else.
The feature that takes three days longer than expected.
The moment you discover that the way you originally designed something does not actually make sense.
That is where theory becomes judgment.
And judgment is what people are usually trying to build when they say they want to become better engineers.
There Is More Than One Good Career
I have also realised how easily we can make young people feel like there is one correct version of success.
Start a company.
Get a job at a famous company.
Become an AI engineer.
Move abroad.
Raise money.
Build a product.
All of those can be good paths.
None of them are mandatory.
You can become an excellent individual contributor and build a meaningful career.
You can become an engineering manager or product manager if you enjoy working between technology, people and business.
You can specialise deeply in infrastructure, security, data, frontend engineering or any number of areas without needing to become a founder.
You can start a company if you are drawn to ownership, uncertainty and responsibility.
But it helps to be honest about what each path requires.
Entrepreneurship may eventually give you greater control over your time, but it can demand an enormous amount of your time before it gives anything back.
Technical work can offer depth, stability and satisfaction, but becoming excellent at it still requires understanding the problem behind the code.
Leadership may give you more influence, but often means spending less time doing the work that originally made you good enough to lead.
There is no path without tradeoffs.
And there are very few permanent decisions.
My own career has moved through different technologies, industries, technical roles and eventually business ownership. I could not have planned every transition when I was starting out.
Some things only became visible after taking the previous step.
The goal is not to copy someone else’s life.
The goal is to understand yourself well enough to choose your tradeoffs deliberately.
Good Mentoring Is Probably Mostly Listening
I used to think mentoring was mainly about sharing knowledge.
It is that, sometimes.
There are technical questions with technical answers.
But increasingly, I think mentoring is about listening carefully enough to understand what someone is actually asking.
A student might ask whether they should learn Python.
What they may really be asking is whether they are behind.
Someone might ask how to start a business.
What they may really be asking is whether they are allowed to want something beyond employment.
Someone might ask whether they need to learn AI.
What they may really be asking is whether everything they have spent years learning is about to become irrelevant.
Someone might ask how to build a project.
What they may really need is permission to start before they feel ready.
The technical question is often only the surface.
This has probably been one of the biggest lessons for me through KamiLimu.
You can give someone technically correct advice and still fail to answer their question.
Sometimes the real work is helping someone find enough clarity to take the next step.
Not the entire journey.
Just the next step.
Teaching Makes You Revisit Your Own Assumptions
There is another benefit to mentoring.
It forces me to examine ideas I have carried for years.
When someone asks why I believe fundamentals matter, I have to explain it clearly.
When someone asks why I chose entrepreneurship, I have to separate the parts of my story that are useful from the parts that are simply personal.
When someone asks whether AI will replace software engineers, I have to go beyond whatever opinion is fashionable that week and think carefully about what I have actually observed changing in my own work.
When someone asks how to become confident, I have to remember that confidence was not something I was given.
It was built through repetition.
Through work.
Through getting things wrong.
Through shipping things that were imperfect.
Through encountering problems I did not know how to solve and eventually solving them.
Through being a beginner more times than I expected.
That is something mentoring and reinvention have unexpectedly had in common for me.
Both require a certain amount of humility.
Teaching has a way of revealing whether you truly understand something or whether you have only become familiar with the language around it.
If you cannot make an idea useful to someone encountering it for the first time, perhaps you do not understand it quite as well as you thought.
That is useful.
What I Hope Students Leave With
I do not expect every student I speak to to follow my path.
They should not.
If mentoring became an exercise in producing younger versions of ourselves, I think we would have missed the point.
Their careers will unfold in a different technological environment from the one I entered.
The tools will be different.
Some jobs will disappear.
New ones will emerge.
AI will almost certainly change the shape of their work in ways we cannot fully predict yet.
But there are a few things I hope remain useful regardless:
- Learn the fundamentals deeply.
- Build things that you care about.
- Do not confuse a tool with an identity.
- Learn how to explain what you are building and why.
- Stay curious about the world beyond your immediate field.
- Get comfortable being a beginner.
- Understand that careers are built through repeated decisions, not one perfect decision.
- Do not wait until you feel fully ready to start.
Most of all, I hope they leave believing that their current position is not their final position.
You can begin with uncertainty.
You can begin without the perfect plan.
You can begin without knowing which programming language you will still be using five years from now.
You can begin while still figuring out who you want to become.
You just have to keep moving.
Perhaps that is what I value most about my time with KamiLimu.
I may be there to help students become better engineers and navigate the beginning of their careers.
But the conversations keep forcing me to examine my own career, my own assumptions and the things I believe about learning.
Mentoring, it turns out, goes both ways.
More from this series
Continue reading the Building Better series