Simply Wrong

I recently saw a social media post where the author had asked an AI what good learning design looks like. It gave them chunking, spaced repetition, varied modality, practice time, and coaching opportunities. Five bullet points. Problem solved. Pack up, everyone; we’re done here.

Note: Bear with me; this isn’t a completely insane rant of an article, though I’ll admit it’s not far off.

Except, of course, we’re not, because the distance between a bullet-pointed list and a functioning intervention inside a real organisation is roughly the same as the distance between a recipe and a restaurant. One is information; the other is years of craft, compromise, politics, budgets, people who don’t reply to emails, and the chaos of trying to change how human beings behave at work.

I want to be clear: I’m all for high standards. I’ve spent a good portion of my writing life calling out weak practice, challenging myths, and arguing that we should hold ourselves to account. I believe that wholeheartedly, and I’m not about to stop. But there’s a growing trend in our space that conflates holding high standards with pretending that the work itself is simple, and that the only reason things go wrong is that we’re collectively incompetent. That framing is dishonest and, worse, it’s unhelpful.

Note: There are areas where I do think we can be harsh on ourselves. Measuring the effectiveness of our work is probably one of the bigger ones, but I don’t think we should fall into the trap of extending the same level of criticism across all areas of our work.

Workplace learning and performance enablement would be straightforward if it weren’t for the one variable that makes it endlessly, beautifully, frustratingly complex: people. Every person in every organisation brings a different history, a different set of assumptions, a different context. We can’t build entirely personalised interventions for every individual, and there’s growing evidence that over-personalisation carries its own risks, but we also can’t pretend that a generic five-step process will survive contact with any real workplace.

The AI element makes this worse, not because the technology is bad, but because it creates an illusion of competence. An LLM can produce a confident-sounding list of best practices in seconds, but it has never had to implement any of them. It has never navigated a stakeholder who changed the brief three days before launch, never tried to measure the business impact of a mentoring programme, never sat in a room where the budget was halved, and the timeline doubled. It doesn’t understand what it produces; it guesses words. And yet, when people see that output, they read into it confirmation that this work is easy and we’re just not doing it well enough.

So rather than leave this as a complaint, here are three things I think we can do to hold ourselves to a high standard whilst remaining honest about the world we work in.

Assume positive intent. When you look at a peer’s work and think you’d have done it differently, remember that every project carries invisible constraints. Budgets, politics, timelines, shifting priorities. Start from the assumption that the person was trying to do good work, because they almost certainly were.

Seek out your own wrongness. Spend more time looking for evidence that contradicts what you believe. Find people who work in contexts you’ve never encountered and discover how their version of best practice differs from yours. Almost nothing in our field is universal, and the sooner we accept that, the better our work becomes.

Lift others up, and grow in the process. Share what you know, write about it, talk about it, debate it over coffee or in a comments section. Every time I’ve tried to help someone else understand something, I’ve exposed gaps in my own thinking and opened myself up to perspectives I hadn’t considered. A rising tide, as they say.

We’re not perfect. We have plenty to improve. But this work isn’t simple, and pretending otherwise helps nobody.