Give people room to be bad at this

Learning AI in public can feel like being asked to perform a magic trick while someone evaluates your career prospects. This issue offers a more appealing starting point: room to try something ambitious, make a mess, and understand what happened. We like learning environments where an imperfect attempt gives you something to work with. Nobody develops judgment by watching the instructor nail the demo.

Citi lets its people team bring the ideas

HR Brew reports that Citi’s HR hackathon brought roughly 180 participants onto 35 teams, with broad challenges and room to propose their own projects.

Our take: Letting people choose a problem gives their existing expertise somewhere to go. The revealing detail is what followed: participants also surfaced smaller everyday uses and became people colleagues could ask for help. We would make room for that quieter contribution alongside the prize-winning project. A colleague who can explain a useful trick is an adoption asset.

AI can help you rehearse before it helps you deliver

Melissa Loble writes in Fast Company about retrieval practice, AI-supported rehearsal, and reflecting on your reasoning as habits that support learning.

Our take: A chatbot that writes your presentation can leave you with polished slides and a fragile grasp of the argument. Ask it to play the audience, and you get a chance to find the weak spot yourself. We’d put more of that rehearsal into AI training. The awkward answer you improve can teach you more than the perfect answer you paste.

One useful move

Practicing for a presentation? Give your usual AI tool the main point and the audience, then ask:

Presentation topic and materials: {{presentation_materials}} Act as a thoughtful audience member. Ask me one question about this presentation that I should be ready to answer. Wait for my reply, then help me make the answer clearer and more specific.

Say your answer out loud before typing it. Keep the improved version in your speaker notes.

Something fun

Give the chatbot a tougher speaking assignment:

You are a museum guide explaining a mysterious artifact called “the office printer” to visitors in the year 2400. Write a short, confidently incorrect tour commentary. Include one theory about why humans kept feeding it paper despite its obvious hostility.

We suspect ritual appeasement. The paper jams were a test of devotion.

A question for you

If your workplace held an award for the most useful failed AI attempt, what would you enter, and what would the trophy look like?