Human-Machine Peer Learning / reciprocal learning

Machine Morality / Axiometry of Small Language Models

speech-to-text, text-to-speech, singing voice generation, computer vision, computational rhetorics etc.

teacher Avatarization / cybridization

Sustainable AI

vibe-coding / vibe-tuning

embedded AI / embooked artefacts

AI-Literacy courses (teaching-about-AI)

Teaching-with-AA is one among many didactic methodologies whereby PRACTICE of human teacher is consciously and intentionally EXTENDED, ENRICHED and COMPLEMENTED *** by artificial assistant (AA) systems.

** but not replaced !!!

Reflect on three main questions of the teaching practice: (What | Why | Whom) do I teach ?

Reflect whether AI assistance / extension / enrichment is necessary. If not, don't use it.

Translate Your pedagogical (resp. andragogical) intent into exact, to-the-point, clear, unambigous INSTRUCTION / prompt composed of exact, to-the-point, clear, unambigous TERMS.

Deploy tools of Your own choice and, ideally, making.

Chisel Your style.

for Winter Semester (indoor) courses I need cca 10 - 20 new foliae every single week

folio - a basic unit of my Open Educational Resources (OER) - is composed of mandatory (title and content) and optional (image, audio, video or fulltext) elements

each folio is just a SUPPLEMENT to teacher's narration

main didactic approach: "moderate multimodality"

LLMs (e.g. GPT5.1 or own model / LoRA) assist by transforming the human-originating andragogical INTENT into densified HTML content fragments

text-to-image (e.g. DALL-E or own model / LoRA) translate the textual content into visual modality

reciprocal creativity: human teacher provides input for the AI, AI assists but sometimes also provides services akin to ΜΟΥΣΑ / orāculum

role of the human teacher: intention, course structuration, curation and, most importantly, awakening inspiration by personalized, HERE & NOW narration

1. I will now upload photos of the attendance list of my students. It has three columns (name, study program, email address). Transform it into a CSV file or a table I can then use as an Excel sheet.
2. Now I uploaded the second attendance list from the second session. If the name already occurred (e.g. Levenshtein distance < 4 ) on the first one then mark the attendance as double attendance. Otherwise, it's only one attendance. Ultimately, create a CSV with two columns for attendance merging the first and second list.
3. Question: Is this GDPR-compliant ? Or should one use mistral-ocr ? Or only Tesseract ?