How I learned AI – my personal map

Over the past two months, several people have asked me the same questions: "Monika, where do you even start with AI?", "Is there a course you'd recommend?", "How did you actually learn this stuff?"

So I decided to write it all down in one place. This is where things stand as of March 2026 – AI moves fast, and some of this might look different a year from now.

0. Where I'm coming from

I spent many years working in product management – at corporations and startups, in Poland and abroad. From 2019, I was part of a startup whose core mission was building AI for healthcare. That was my first real, deep encounter with AI – from the inside, in practice, in a context where it genuinely affected people's lives.

That's the starting point.

1. Data science – deep, intense, and probably not for everyone

In 2024, I completed a 17-week data science bootcamp at WBS Coding School in Berlin (680 hours, September 2024 – January 2025). The curriculum covered data analysis in Python, SQL, machine learning, data storytelling in Tableau, cloud computing, and generative AI.

Would I recommend this to everyone who wants to get into AI? Honestly – no. That level of technical depth isn't necessary if your goal is to use AI at work or implement it in an organisation. Unless you want to go deep into the technology itself – then absolutely.

What it gave me was something I didn't expect: empathy for models, and a real understanding of why data quality matters so much. I know where errors come from, why AI "halluccinates," and what it simply can't do. And that – as it turns out – is incredibly useful in conversations with clients and workshop participants.

2. Allie K. Miller – AI from a practical, actionable angle

Then I came across Allie K. Miller and her courses. Her flagship programme is The AI-First Academy, but there's a lot more on her website. I follow her on LinkedIn and Instagram too.

Her expertise genuinely impresses me – she approaches AI from a very practical, ready-to-implement angle. She cuts through the noise and focuses on what actually works. And the fact that she holds such a strong position in the AI world while remaining so accessible – that means something.

3. AIDEAS and Umiejętności Jutra – for those who want to get oriented

When I moved back to Poland, I took part in both Umiejętności Jutra (Google's "Skills of Tomorrow" programme) and AIDEAS – and I'd recommend both.

Not because they'll teach you to deploy AI across your entire company overnight – but because they help you map the terrain before you start investing time and money in specific tools or courses. Both programmes are free, and there's a new cohort starting now, so worth checking out.

4. Using tools every day – this is where I learn the most

No course has taught me as much as simply using tools every day. I test, observe, and compare – ChatGPT, Gemini, Claude. I track what Google is building across its ecosystem, how models are evolving, what's new each week. Good newsletters, podcasts, and tutorials make a real difference.

One tool stands out for me above the rest: Claude by Anthropic.

5. Why Claude?

It's not just that Claude is the best model I've found for writing, thinking, and working with documents. It's what stands behind it. Anthropic – the company that builds Claude – is one of the very few players in this space that treats AI ethics and safety as a foundation, not an afterthought.

Claude Code and Cowork – why this changes how managers work

To understand this, you need to take a step back. For years, AI was a chatbot – you ask, it answers, that's it. Claude Code was built for developers, as a natural, conversational way to write and develop code. But it quickly turned out that many non-developers started using it too. That was a clear signal: if this works for engineers, it can work for everyone.

Claude Cowork is the answer to that signal – a more accessible version of the same idea, designed for every knowledge worker in an organisation. You give it access to your files: documents, spreadsheets, Google Drive folders, emails, the systems you already use. It reads them, understands the context of your organisation, and executes multi-step tasks from there – gathering data from multiple sources, analysing it, writing a report, building a presentation – without needing you to re-explain everything at each step.

Companies can build their own versions of Claude tailored to their processes and institutional knowledge. A "Claude for HR" knows your onboarding flow and contract templates; a "Claude for Finance" understands your models and reporting structures. This isn't a generic chatbot – it's a tool that operates within the specific context of your organisation.

6. BRAVE – structured learning in Polish

I've been following BRAVE and their course ecosystem closely. They have programmes for different profiles – product people, developers, sales, marketing, and managers.

I'm likely joining AI Managers myself starting in May. Both the instructors and the curriculum genuinely convince me. My direction is strategic and organisational, not deeply technical – and that's exactly what this course is about.

7. What if you want to implement AI in your organisation?

I work with companies in different ways. Sometimes it's a two-day workshop to map out where AI could genuinely add value and what the options are. But with some organisations, I work in an ongoing series of sessions – and that consistently delivers much better results, because there's time to actually get comfortable with the tools and respond to real, evolving needs.

Two examples from my practice:

Company – implementation from scratch. I've been working with one company for two months now. We started with prompting, then gradually introduced tools for their specific needs – AI for social media management, content creation, grant writing, and research. That kind of pacing gives people time to genuinely absorb the tools, not just be shown them. Already, I can see a real growth in what they're able to do, and the organisation is now ready to go further – deeper automation, deeper integration – in a sustainable, thoughtful way. Having someone guide the process with them, they're achieving far more than they would on their own.

Researchers and academics – AI with ethics at the centre. This is a very different context, but equally important. I had the pleasure of working in this space with Nicolaus Copernicus University in Toruń. Here, data security and ethics are the central concern – it's about finding tools that support the work without replacing the researcher's judgment. AI is excellent for literature reviews, mapping connections across sources, and broad research tasks. But the line between support and substitution has to be drawn deliberately and kept visible.

Implementing AI in organisations requires someone who understands both the tools and the business processes – and who brings a lot of empathy to the work. Because AI isn't just about new tools or new workflows. It's fundamentally about changing how people think and how an entire organisation operates. And when the change is that significant, what matters most is taking care of the people going through it – not just increasing efficiency, but doing so in a way that's sustainable and responsible. That part is really, really important.

8. The AI Generalist – a profile for the future

There's a lot of talk right now about the rise of the AI Generalist – and I think it's right. These aren't purely technical developers, and they're not traditional managers either. They're people who can bridge product thinking and technical knowledge, who ask good questions, and who know how to translate all of it into real organisational processes. A human presence between the machine and the people around it.

And that brings me to something that rarely gets mentioned in AI conversations.

9. Empathy – for machines and for people

Yes, I'm using the word "empathy" in the context of AI.

Empathy for machines means understanding their limitations, their quirks, why they go wrong, and when they can actually be trusted. Empathy for people is even more important – especially when you're driving change in an organisation, when people are afraid of being replaced, when you need to explain complex things in plain language.

Without that, technology is just a tool. With it, it can be a genuine transformation.

10. My "and" – technology and wellbeing

I won't pretend my interests stop at technology. I'm a certified Forest Bathing Guide and I practise breathwork, yoga, and meditation. I genuinely feel that the further we go into AI and technological development as a society, the more we'll need to come back to what's most human and most natural. Getting caught up in technology and the constant stream of new things limits us – as people.

What I want to build is exactly that combination: technology and wellbeing. For me, these two worlds aren't in conflict – they're inseparable, and they can reinforce each other in a profound way, both for productivity and for a sense of calm and meaning.

Technological development is inevitable. But how we develop it depends on who we are as people. The more we return to nature, to breath, to stillness – to the things that aren't technological – the more ethically and responsibly we'll be able to develop the technology itself. And that, I think, is the heart of all of this.

Summary

If you're starting out:

• AIDEAS and Umiejętności Jutra – free, to get oriented

• Allie K. Miller – AI from a practical, actionable angle

• BRAVE – if you want structured development in a specific direction

• Claude, ChatGPT, Gemini – use them, test them, pay attention every day

If you're implementing AI in an organisation:

• Find someone who understands both the business and the tools

• Be sceptical of promises about "10 agents in 3 hours"

• Commit to a process and to taking care of people – not a one-off training

This is far from a complete picture – there are many more sources, courses, and tools out there. These are simply the ones that have worked for me. I also learn a huge amount from conversations with other practitioners, clients, and workshop participants, and from attending conferences where you can see first-hand where all of this is heading.

AI is a marathon, not a sprint. Good guidance and patience matter more than the best course money can buy.

Monika Kotus – AI consultant & trainer | monikakotus.com