Just like everyone else in the UK we are getting very excited about National Tea Day which takes place on the 21st April. The day is a great opportunity to share tea brewing preferences, and the strength of the perfect 'cuppa' is always hotly debated. So we though it would be interesting to get the view of AI.
This morning I went for a routine dental check‑up. What struck me was how naturally my dentist fell into a comparative frame. He wasn’t mentally ticking boxes on a rubric. He was looking at two artefacts and asking, “Which represents a healthier, more stable situation?”
The story of the Barometer Question is usually told as a joke at the expense of an inflexible examiner. But if you look at it through an assessment lens, it is really a story about design failure, construct clarity, and the importance of a strong holistic statement of quality.
Economist Charles Goodhart warned that “when a measure becomes a target, it ceases to be a good measure.” The paperclip AI is that warning turned into a toy: once “number of paperclips” becomes the only thing that counts, every other value is expendable. When I look at how AI is being introduced into education, especially assessment, I sometimes worry we are drifting towards our own paperclip machines.
All RM Compare sessions can allow you to extract very detailed data sets (Advanced and Enterprise Plans). We already provide you with some great reporting, however the latest Generative AI tools might be able to give you even greater insights.
In this series, we’ve been arguing that sovereignty is the only sustainable path for professional judgment in an agentic world. We started by looking at the infrastructure we’ll need for sovereign intelligence, then at the difference between sovereign systems and AI wrappers, and most recently at dark patterns that steer people into feeding someone else’s models.
In our series on digital sovereignty, we have explored how "Harvesting Models" use dark patterns to bypass consent and how "AI Wrappers" quietly extract value from your institutional judgment data. But there is one final, definitive test of a vendor’s intent: What happens when you want to leave?
In earlier posts we argued that sovereignty is the only sustainable path for professional judgment, and that organisations need infrastructure which lets them own and operate their own intelligence. There’s another layer to this: even if the architecture is centralised, the interface can quietly decide who wins. The way AI is presented in a product can make the difference between a conscious choice and a gentle shove.
In the last two posts, we argued that sovereignty is the only sustainable path for professional judgment, and that the real work now is building the infrastructure that lets organisations own and operate their own intelligence.
In 2011, Marc Andreessen famously noted that "Software is eating the world." By 2024, Nvidia’s Jensen Huang updated that reality: "AI is eating software." As we move into 2026, we are entering the final stage of this evolution: Agents are eating the enterprise.
The vision of a truly data-driven system is no longer a distant goal; it is the immediate frontier. In 2026, we are witnessing a fundamental shift in how organisations understand data and expertise. This isn't just a story for schools; it is a global movement toward Sovereign Intelligence that spans every sector where human judgment matters.
The recent publication of the Data and Information to Support Learning and Improvement Impact Assessment by the Welsh Government marks a watershed moment for education in Wales. It signals a definitive move away from narrow, high-stakes accountability and toward a holistic, learner-centered "Information Ecosystem."
In the world of education and assessment, the terrain shifts daily. A “map” - whether it’s a static dashboard or a rigid three-year product plan - is only truly useful if the world stands still. But when the environment moves faster than your plan, a map quickly becomes a liability.