The Night I Realized I Wasn’t an AI Expert
It was 2 a.m. in Dubai, over the weekend, in early 2020, and I was still watching yet another “Deep Learning in 10 Minutes” video, as most of us do to open YouTube, bookmarked dozens of courses, and got lost in a sea of buzzwords.
On this specific night, on my home office desk, a cold tea, an open Coursera tab, and three notebooks filled with equations that made less sense with every reread.
I had spent weeks jumping from TensorFlow tutorials to blog posts about machine learning algorithms.
Everyone online sounded like a genius.
And there I was, with a decade of experience in enterprise tech, feeling like I’d just entered kindergarten again.
That night, I closed the laptop and wrote a single line on a sticky note:
“AI is not about knowing everything - it’s about knowing where to start.”
That moment changed my entire approach to learning, and I have already written and recorded a podcast about how I started an MSc in Artificial Intelligence as a formal program, not to get stuck on where to start, but to follow a structured path.
Today, after years of working with clients on AI adoption and Generative AI strategies, I often meet people who feel exactly how I did.
So here are the five things I wish someone had told me when I was starting, and the five similar things I wrote in 2016 when I was developing PaaS "5 things I wish I’d known when I started developing PaaS Extensions for Oracle SaaS Applications", in the early days when organizations were moving from on-prem applications to SaaS.
1 - Start with a Problem, Not a Model
Early on, as a beginner, I chased models, CNNs, RNNs, Transformers, etc, etc.
Real progress began only when I picked one business problem to solve.
“The best AI projects start as business questions, not Kaggle competitions. Although Kaggle competitions give you the learning depth & breadth
Try this: Pick one routine frustration, like manual report writing, or ask the business users, and automate it.
The learning sticks without any hype when you see a result that matters to you or your team.
2 - The AI Landscape Moves Fast, But Foundations Don’t
Now, after almost 3 years into the LLM era, every year brings new tools, yet the core never changes, and it is the same as it was when we were in the Digital Transformation Era:
Data - Pattern - Prediction - Action
If you start understanding early how data becomes decisions, you’ll never fall behind.
Focus less on chasing models, more on connecting: inputs - insights - impact.
3 - Tools Change, Mindset Lasts
I am recalling early days, and I can say that I wasted months debating whether to learn TensorFlow or PyTorch.
Later, I realized the tool doesn’t define the learner; habits do, and I learned how to form a habit from the book 'Atomic Habits'
Three that shaped my career:
- Curiosity: experiment even when unsure.
- Reflection: Write down after every experiment, and that's the reason I have been writing blogs and articles since 2009
- Documentation: record results like a mini case study, write blogs and articles, and my writing routine pushes me to refine whatever I am doing.
The real AI differentiator is not code but clarity.
4 - Your Career Will Evolve Around AI, Plan for It
The way progress is happening on the AI front, with significant investments pouring into AI Infra development, and even non-technical roles are now being started, including AI-powered workflows.
In my day-to-day interactions with customers, friends, and colleagues, I have seen several departments quietly adopt AI to automate different processes.
Those who learned early became AI champions in their departments, not coders, but translators between tech and business.
That’s where future leadership lies.
5 - Community and Consistency Beat Solo Learning
Learning AI can feel an isolating action.
When I started the AI Tech Circle community, it wasn’t to teach; it was to learn with others.
Sharing small weekly actions built momentum and friendships that still inspire me today.
If you only do one thing this month, post your first AI experiment.
Tell the world what you tried, even if it failed.
Momentum beats perfection.
Takeaway Framework - The 3-E Rule
Explore: Try new tools & concepts for 30 days, the outcome will be to build the Curiosity muscle.
Experiment: Build tiny use cases in your own workflow, the outcome will be to build Practical confidence
Expand: Share, teach, and collaborate, the outcome will be to build visibility & career growth.
If you’re starting in AI, don’t try to learn everything.
Pick one problem, one tool, and one community. That’s enough to start your momentum.
A few references and further reading material for you:
