AI Advent 2025; 25 Days of Learning
🎄 AI Advent 2025: 25 Days of AI Insights
Welcome to AI Advent 2025 — a daily series of short, practical reflections on
Artificial Intelligence in research, education, and data-intensive science.
Each day in December, a new post will be published. You can read the days
in any order and return at any time.
- Day 1 – AI is not magic
- Day 2 – When NOT to use AI
- Day 3 – Trustworthy AI in science
- Day 4 – Bias in scientific datasets
- Day 5 – AI and HPC: why scale matters
- Day 6 – Explainability vs performance
- Day 7 – AI hallucinations in research
- Day 8 – AI in astronomy: lessons learned
- Day 9 – Reproducibility in AI pipelines
- Day 10 – Open science and AI
- Day 11 – Energy cost of AI models
- Day 12 – Human-in-the-loop systems
- Day 13 – Ethics is not a checkbox
- Day 14 – Data quality beats model complexity
- Day 15 – Foundation models for science
- Day 16 – Generalisation vs memorisation
- Day 17 – AI and peer review
- Day 18 – Automation vs insight
- Day 19 – Teaching with AI responsibly
- Day 20 – AI governance in universities
- Day 21 – Learning from AI failures
- Day 22 – AI beyond the hype
- Day 23 – Measuring impact, not novelty
- Day 24 – What students should learn about AI
- Day 25 – The future of AI in science