A new ladder

Dr Phil Hardman’s latest analysis is fascinating: in her sample of 379 UK/US L&D job ads (June–August 2026), only 1% were junior roles. That’s because AI now drafts lessons, quizzes and videos. Humans are hired – if at all – to specify, judge and measure. Outside L&D, the same is happening – PwC’s 2026 barometer… Read More A new ladder

Evidence first

Over the last few weeks i’ve been reading. I’ve been reading loads, some good and some poor. A hint to people if you’re going to write a book about AI: If you’re referencing historical estimates of development time for eLearning, estimates of unused or unapplied learning, or forecasts about AI token consumption, cite your sources.… Read More Evidence first

Andrew Bot

I decided, after last week’s post, to delegate all my posting to an AI agent. This is how the conversation went: Andrew: Write me a LinkedIn post about resilience. Andrew Bot: Here’s your post.“Resilience isn’t just bouncing back. It’s bouncing forward.”I used three em dashes, opened with “Here’s the thing” and described resilience as a… Read More Andrew Bot

Average engine

We seem to forget that the purpose of Large Language Models is to predict the most likely next word, phrase, or sentence. Effectively, that means using the tool to generate content that will, by its design, move you towards the mean. Remember that the tool is an average engine and you’re better placed to understand… Read More Average engine