Kerstin Cable On AI, Habits And How To Keep Going As A Language Learner

I've known Kerstin Cable for a long time – she was one of my very first internet friends when I started out online, back when we were both trying to build something in the language learning space.

She's been thinking seriously about how people learn languages ever since, and she does it with the kind of practical wisdom that only comes from years of teaching real learners.

Kerstin is the founder of Fluent Language and the creator of AI Language Club, and when we chatted, I wanted to hear her take on two things she's thought about more carefully than almost anyone: how AI fits into a language learner's toolkit, and how to build habits that actually last.

If you prefer listening to reading, hit play on the podcast version of this article below or on the YouTube version at the top of the page. Otherwise, here's what you'll discover in this post:

Who Is Kerstin Cable?

Kerstin Cable is originally from Germany and has been running fluent language.co.uk since 2012.

She started out as a German tutor but quickly found she had strong opinions about language learning more broadly – about the psychology of it, the habits that sustain it and the traps that derail it.

She's also the teacher behind the German Uncovered beginner course, which we built together, and one of the most thoughtful people I know in this space.

Her language CV is impressive. Despite getting C's and D's in French at school, she stuck with it through university and has been a fluent French speaker for years – most recently cycling around northern France and finding she could get around with ease.

She has decent Spanish, has dabbled in seven or eight other languages and, for the last decade, her main target language has been Welsh. She has now reached what's described as “past advanced” level, which in any language is a serious achievement, and in Welsh it's remarkable.

When I asked her what had kept her going through all of it, she gave me an answer that I think gets to something important: “I seem to have the gift of enjoying being crap at stuff.” That willingness to sit with being a beginner, she thinks, is more valuable than any method.

AI As A Practice Tool – Not A Replacement For A Teacher

I've been cautious about AI at StoryLearning, mostly because our method is built on fundamentals that have always worked and I'm wary of disrupting something that isn't broken.

But I'm increasingly curious about how others are using it, so I asked Kerstin for her honest high-level view.

She was immediately enthusiastic, but also precise about where AI belongs. For her, it's not a rival to Duolingo or Memrise — it's a step change beyond them.

“It simulates natural language. It's just so much richer than your Duolingos, your Memrises and all those kinds of apps,” she said.

The key word she kept returning to was practice. AI, in her view, is the most versatile language practice tool that learners have ever had access to.

What it isn't, she was equally clear, is a replacement for a teacher. When our mutual friend Gareth from How to Get Fluent suggested that's where AI might be heading, her response was that the thought hadn't even occurred to her.

“We learn languages for connection. We want the human connection, and a good teacher will challenge you, but in a safe way.”

The teacher and the AI tool serve entirely different purposes — one is about guidance, relationship and accountability; the other is about access, availability and practice volume.

What Can You Actually Do With AI?

This is where the conversation got practical. Rather than offering a single progressive AI curriculum, Kerstin's AI Language Club is built around short, specific techniques – bite-sized ideas you can plug into what you're already doing.

She gave me two examples that I found genuinely useful.

The first is vocabulary preparation. You go into ChatGPT or Gemini, tell it your level and your context — “I'm an advanced Welsh learner and I'm about to visit a chip shop” — and ask it to generate the specific vocabulary you need.

Kerstin mentioned a professional interpreter who uses exactly this approach before EU conferences: give me the 25 words I absolutely must know for tomorrow's session on agricultural machinery.

It's not revolutionary in concept, but having that level of instant, personalised preparation available for free is something genuinely new.

The second technique she calls the Transformer. You ask the AI to have a conversation with you in your target language, but with a twist: for everything you write, it offers three different ways to say the same thing before continuing the exchange.

You say something simple – “the weather is nice today” – and it reflects back three natural alternatives before the conversation carries on.

The idea is to expand your expressive range incrementally, without overwhelming you or pulling you out of the flow. “It stretches you a little bit,” she said, “and then you move on. You still have the conversation.”

Both of these replicate what a good teacher does naturally: the vocabulary preparation is focused research; the three alternatives is scaffolding. The difference is the access. You can do it at eleven o'clock at night, for free, in whatever language you're working on.

Does More Choice Make Language Learning Harder?

This is a tension I've felt myself, and I raised it with Kerstin directly. I told her about learning French in the year 2000, when the only place I could find a decent language book was a specialist shop in Soho.

I'd browse eight titles, pick one and that was my book for the next three months. There was something clarifying about that constraint.

Contrast it with now, when I have unlimited access to audio, video, podcasts and AI – and I often feel less focused, not more.

Kerstin agreed that decision fatigue is real, and she connected it to something she'd read about in the book Modern Romance – the observation that having too many options in dating leads to chronic dissatisfaction, always wondering if someone better might be out there.

The same pattern seeps into language learning: you're always wondering whether the next app, method or tool might be more effective than the one in front of you.

But she pushed back on the idea that AI is responsible for this. “This was a problem when YouTube arrived,” she said. The abundance isn't the issue – the issue is learning to trust your own judgement about what's working.

Her advice is to be genuinely observant about what feels satisfying in your learning, and to allow yourself to experiment without anxiety.

“It is perfectly okay to try asking ChatGPT to transform your sentences this week and then not do it next week. Just naturally allow what sticks to stick.”

How To Build A Language Habit

Kerstin Cable has been thinking about habit formation for years – it's the foundation of her Language Habit System – and her approach is refreshingly practical and unsentimental.

Her core concern with streak-based apps is that they train you to be good at using the app rather than good at the language.

The goal she prefers is what she calls daily contact: some form of engagement with your target language every single day, however small.

Some days that's a formal lesson; other days it's having the Welsh radio on while you make breakfast. “What is the minimum I can do?” is the question she encourages learners to ask.

We also talked about activity pairing – attaching your language habit to something you already do reliably. When you get dressed, put on a YouTube video. When you're in the car, the podcast goes on without discussion.

She mentioned something I hadn't thought about before too: decorating your physical environment with the language. She has Welsh art on her walls at home as a quiet daily signal that Welsh is part of her life. Small things, but cumulatively they reinforce a sense of identity as a speaker of that language.

Which brings her to what she thinks is the most important ingredient of all. Accountability through another person – a teacher, a language partner, a class – works well for many learners, and she sees its value clearly.

But even more durable, she thinks, is having the identity of a learner. “Being a learner is the most wonderful feeling and experience you can have,” she told me, and she meant it without irony.

As long as you can hold onto that – the genuine pleasure of not yet knowing, and moving incrementally towards knowing – you'll find a way to keep going.

The Learner Who Keeps Showing Up

What I took away from my conversation with Kerstin Cable was less a set of techniques – though the techniques are good – and more a way of thinking.

AI is a powerful practice tool, and learners who ignore it are leaving something genuinely useful on the table. Daily contact matters more than intensity. Identity matters more than method.

And the choice paralysis that comes with having too many options is a real obstacle, one that requires self-awareness rather than more resources to overcome.

Kerstin has been helping learners navigate all of this for well over a decade, and what strikes me most about her approach is its consistency: she's never been interested in finding the perfect method, only in helping people stay in the game long enough to get good.

She talks about enjoying being crap at stuff as though it's just a personality quirk, but I suspect it's actually the secret. The learner who keeps showing up, even when it's difficult, even when progress is slow, is always going to outrun the one who's still searching for the perfect shortcut.

Olly Richards

Creator of the StoryLearning® Method

Olly Richards is a renowned polyglot and language learning expert with over 15 years of experience teaching millions through his innovative StoryLearning® method. He is the creator of StoryLearning, one of the world's largest language learning blogs with 500,000+ monthly readers.

Olly has authored 30+ language learning books and courses, including the bestselling "Short Stories" series published by Teach Yourself.

When not developing new teaching methods, Richards practices what he preaches—he speaks 8 languages fluently and continues learning new ones through his own methodology.