Kerr reviews Moorhouse & Wong’s Cambridge Element Generative AI and Language Teaching (2025) as promotional: thin on practice, selective on evidence, and built on the unexamined assumption that GenAI makes teaching more efficient. ‘AI for Good’ framing echoes a UN initiative sponsored by big vendors.
Sources cited
Bender & Hanna (2025) The AI Con; Selwyn, Ljungqvist & Sonesson (2025) on teachers’ work around GenAI’s frailties; OECD (2023) Generative AI in the Classroom: From Hype to Reality; Adams (2025).
Implications
- Ask for evidence behind “AI saves teacher time” (e.g., grading) and test it in your context.
- Check who sponsors a “for good” event or resource.
- For my platforms: show learning outcomes, not just workload savings; make AI decisions visible (why an answer was marked).
Counterpoint
Harmer (comment): the evidence gap is real, but AI helps in practice; train critically aware students. Kerr: critical-thinking training has little evidence outside narrow domains.
Connections
Solutionism - define the problem before the technology · Critical AI literacy is hard to teach · Does ChatGPT enhance learning - performance is not learning