---

A good teacher can explain something brilliantly and create a space in which learning can happen. A good tutor can do the same one-to-one and give careful, individual feedback. Then the lesson ends. Sometimes the learner is still sitting there with a draft open, a task in front of them and one very simple question:

What do I do next?

Episode 111 of iAntonio Media – Living the Language begins in that space between lessons, but the conversation does not stay there for long. It moves into technological change, professional adaptation, critical thinking, the difficulty of showing what we have actually learned, and a question that has followed me through several very different parts of my working life:

When the tool changes, what part of my contribution becomes more valuable?

That question eventually brings me back to education – and to Sybille.

Learning does not stop when the appointment ends

Anyone who has taught for a while will recognise the situation. Something appears to make complete sense during the lesson. Twenty minutes later, the learner is struggling to apply it independently.

Sometimes a student can repeat an explanation perfectly without really knowing what to do with it. In the episode I compare this, with some affection, to the parrots I had as a child: they could say hello and good morning beautifully. Whether they understood the social significance of either was another question entirely.

That little distinction becomes quite important.

Being able to repeat something is not necessarily the same as being able to use it. And being able to use something once, with help, does not necessarily tell us what the learner will be able to do later, independently or in a different situation. For teachers and tutors, that creates a practical limitation. Learning often needs repetition, another attempt, another explanation, another round of feedback or simply more time.

Human attention is finite. So is the school timetable. So is the amount of individual feedback one teacher can give to several learners while still doing everything else the profession requires. The lesson can end at four o’clock. The learner’s question does not necessarily end with it.

I have seen technological disruption before

Part of the reason I am relatively calm about the current technological shift is that it is not the first one I have experienced personally. I have worked in bookkeeping. I trained and worked as a quantity surveyor. I studied business administration. I taught myself photography and worked professionally and exhibited as a photographer. I have now spent almost two decades teaching and training.

Across those different fields, technology has repeatedly changed the work. Sometimes it made a task faster. Sometimes cheaper. Sometimes easier. Sometimes a skill that once required specialised equipment or knowledge simply became much less scarce.

I remember bookkeeping before the spreadsheet had absorbed so much of the routine calculation. I remember the enormous amount of manual measurement, calculation and document handling involved in quantity surveying. I remember entering photography at a moment when some analogue photographers were deeply suspicious of people who had entered through digital photography. And then, rather amusingly, I eventually became one of the people resisting the next shift myself.

For quite a while I rejected smartphone photography. I had invested in cameras, lenses, technique and a particular photographic identity. A phone appeared to be doing computationally what I had learned to achieve through optics, light, equipment and practice. Eventually I changed my mind. These days I am quite comfortable saying that the best camera may simply be the one you have with you – provided you know what you are trying to do with it.

I have not always greeted technological change gracefully. That is part of the point. My interest in Ai does not come from having floated serenely above previous disruptions. It comes partly from having had to adjust to them.

When the tool changes, what becomes more valuable?

Technological change did not make bookkeeping, quantity surveying, photography or teaching disappear. It changed parts of the work. And each time that happens, there is a temptation to concentrate almost entirely on what has become easier for the machine. I am increasingly interested in the other side of that question.

What becomes more valuable because the tool can now do something that used to occupy more of our time?

That question has now arrived very visibly in education. Generative Ai can explain grammar. It can suggest vocabulary. It can reorganise a paragraph. It can correct surface errors. It can produce fluent English in seconds. The existence of that capability does not necessarily tell us very much about what the learner has learned.

The answer improved. Did the learner?

If a machine can produce a more elegant paragraph in three seconds, the quality of that paragraph is only one part of the educational picture. Can the learner explain what changed? Could they make a similar decision themselves? Could they recognise when the suggestion does not fit the task? Could they disagree with it? Could they take an idea that worked here and use it appropriately somewhere else? And if the answer became better, what evidence do we have that the learner became better at anything?

Those questions become particularly interesting in advanced language learning because students are not only expected to know things. They have to show what they can do. In the English Abitur, that may mean moving between different kinds of intellectual work: explaining, summarising, analysing, assessing, comparing or commenting. The language may be fluent while the response still fails to do what the task requires.

A learner may understand a text but struggle to show that understanding convincingly. They may know the difference between personal opinion and evidence in principle yet fail to maintain that distinction in a piece of writing.

They may recognise metaphor, irony, imagery or repetition but still need to explain what the writer is doing with those choices and how they contribute to the text. That movement between knowing, showing, explaining and applying is one of the reasons I remain fascinated by language education. It is also one of the reasons I became interested in building a specialised Ai learning environment rather than simply giving students access to another answer generator.

Where Sybille enters the picture

Sybille started to make sense to me as a feedback space that learners could return to when they reached that familiar point:

I am not quite sure what I should do here.

Or:

I know something is wrong, but I do not know what to change next.

The purpose is deliberately limited. Sybille is not a teacher trapped inside a computer, and she is not an oracle. She supports students preparing for the English Abitur in Berlin and Brandenburg in clearly defined areas of their own work.

Three examples from the episode are especially important.

Focused feedback

A learner can begin with feedback that identifies two genuine strengths and one clear next step. The strengths are there so that useful parts of the work are not accidentally destroyed during revision. The next step gives the learner something concrete to do. The student then has to revise. More detailed feedback can follow, but the first response does not need to become a catalogue of everything that could possibly be changed.

Task context

Fluent English is not enough if the learner is answering the wrong question. Sybille works from the task context: what the learner has been asked to do, which operator is involved, who the intended audience is and what kind of language or register fits the situation. That distinction becomes particularly useful when a text sounds polished but has drifted away from the actual task.

Preflight

Preflight brings the learner back to readiness and ownership before they consider the work finished. Can you explain the choices you made? Why is the argument organised this way?

What changed during revision? Could you defend the wording if your teacher asked you about it? What might need to change if the audience or task changed? The point is not to receive approval from a machine. The point is to create another opportunity to examine the work before handing responsibility back to the learner.

See the Demo

There is a point at which describing Sybille becomes less useful than simply showing what happens. I have therefore published a longer Demo/Tutorial in which I work through one of Sybille’s learning processes and show how the system approaches student work, feedback, task context and ownership. It is deliberately more detailed than a social-media clip. If you want to understand what I mean by responsible Ai use in this learning context, this is the place to look.

Watch the Sybille Demo / Tutorial on TrainingTree.

The second Sybille pilot begins today

This episode is being published on 14 September 2026, the first day of the second Sybille pilot programme.

The pilot runs until 15 December 2026 and is free for selected students in Q1–Q4, Grundkurs or Leistungskurs, who are preparing for the English Abitur in Berlin or Brandenburg.

This is not an anonymous open-access trial. Students or parents contact me first, we have a short personal orientation conversation, and we look at whether Sybille fits the learner’s current situation and whether meaningful participation in the pilot is realistic. Participants are expected to use Sybille over time as part of their own preparation and to provide constructive feedback about their experience.

That feedback is particularly valuable to me now. I have spent a long time building Sybille. The next stage is to see more clearly how learners actually use her: where the design helps, where something remains unclear, which forms of support are useful and what still needs to change.

---

Read the full information about the free Sybille Pilot Programme 2026.

For students and parents in Berlin and Brandenburg (Für Schüler:innen und Eltern in Berlin und Brandenburg)

Sybille ist eine von einem Berliner Englischlehrer entwickelte, KI-gestützte Lernbegleitung für die Vorbereitung auf das Englisch-Abitur.

Das kostenlose Pilotprogramm läuft vom 14. September bis 15. Dezember 2026 und richtet sich an Schüler:innen in Q1–Q4 im Grund- oder Leistungskurs in Berlin und Brandenburg.

Vor einer möglichen Teilnahme findet ein persönliches Orientierungsgespräch statt. Die Nutzung bleibt schülergeführt: Eigene Texte, eigene Entscheidungen und die Verantwortung für das eigene Lernen bleiben bei den Schüler:innen.

Informationen zum kostenlosen Pilotprogramm

Eine wissenschaftlichere Perspektive (KI-Feedback fürs Englisch-Abitur: Verständnis prüfen)

---

One question I am taking with me

This podcast has always been, among other things, a conversation with ideas. I reserve the right to change my mind. So does everybody else.

I do not believe we can stop technological change simply because some of its consequences make us uncomfortable. I also do not believe that every new capability deserves automatic enthusiasm. What I can do is keep asking questions about the kind of learning, judgement and responsibility I want to support.

For Episode 111, the question that stays with me is still the one that appeared near the beginning:

When the tool changes, what part of our contribution becomes more valuable?

And for the learner sitting alone with the draft after the lesson has ended:

What do I do next?

That is the space I am exploring with Sybille.

Sybille on TikTok

Sybille on YouTube

---

---

---

Credits

Written by: Ian Antonio Patterson,

Music: Universal Music (Licensed), Andy Beck,

Post Created with Deliberate AI assistance.

Sybille: Sybille.tech

DISCLAIMER “iAntonio Media”, is a self-sponsored independent podcast brought to you by Ian Antonio Patterson. The show is essentially an art project where we explore different career roles, perspectives and converse with ideas in a non-moralizing, non-prescriptive and non-pretentious way. You get to take from it what you want. It is not exactly made for children and may at times feature sensitive content, based on a wide repertoire of unique lived experiences of people. Listener discretion strongly advised. The show as usual explicitly promises no answers to anything whatsoever. Everyone here speaks for themself, nothing is cast in stone, and we all are, in true human fashion expressly allowed to change our minds. “iAntonio Media” does not offer advice based on the professional opinion of licensed Lawyers, Psychologists, Psychiatrists, Teachers, Moral Authorities, or the like. I am myself a trainer but more likely than not, I am not your trainer… When in doubt consult a professional – and do your own research. For more information, questions, or contributions – feel free to visit @iAntonio_Media at YouTube, Instagram or TikTok, or the official website www.iAntonio.com and use the contact form.

---

iAntonio Media

Thanks for listening, looking forward to hearing from you and bye for now. 

---

Ian Antonio Patterson – ShowHost, Trainer Coach

Sustainable Language Acquisition = Nachhaltiger Spracherwerb.

---

---