GenAI Skills at Work: What Really Needs Training
Generative AI Skills is the second-largest category in the corpus with 967 papers. But most programs train only prompting. What really counts at work goes beyond that.
3D: Konzept-Graph (Knoten & Kanten)
1. The skills that count
- Interpreting outputs: detect hallucinations, check evidence, validate sources.
- Selecting tools: which GenAI tool for which task (RAG, agents, multimodal).
- Ethical judgment: detect bias, protect privacy, take responsibility.
- Integration: embed GenAI into existing workflows (not as an add-on).
2. What NOT to train
Tool operation alone. Definitions emphasizing programming/statistics miss what workers actually need (AI Literacy Assessment Revisited). See Design for the Job.
3. In the ALaaS offer
GenAI skills are part of the team program (Stage 2): role-specific training per the playbook, with objective baseline/exit assessment. See Offer tiers.
Further reading
Sources: Guide 1 (Implementation), §2.2/§4; corpus of 16,459 papers (as of 2026-10). EOF echo "en B done"