The BBC Computer Literacy Project 2.0

I have spent the better part of 40 years as an IT professional, from the days of punch cards on ICL mainframes to architecting agentic frameworks in the age of autonomy. Looking back, there is a specific moment in the early 80s that feels more relevant today than ever: the UK Government-led BBC Computer Literacy Project.

Back then, the mission was simple but radical: introduce the general public to the microcomputer. It was not just for hobbyists. It was a national imperative to ensure the UK was not left behind by a fundamental shift in how the world operated.

Today, we are standing at an even more significant pivot point. But while the 80s were about learning to type and load tapes, 2026 is about understanding the sovereignty of intelligence.

We do not just need a few more prompt engineers. We need a national AI literacy course for schools and the general public. Here is why.

1. Moving beyond the vibe coding era

After using tools like Cursor and Claude for the best part of two years, the biggest lesson has not been about syntax. It has been about communication and structure. Many people today treat AI like a magic black box where you vibe code your way to a result.

As I have noted previously, this is a recipe for failure. Just as the 80s taught us the logic of IF...THEN, today's literacy must teach the architecture of intent.

If we do not teach the public how to articulate precise goals, define constraints, and anticipate risks, they are not using a superpower. They are simply exposed to inference inflation.

2. The identity recalibration

We are witnessing a destabilising shift where identity, long anchored to toil and manual labour, is being challenged by autonomous agents. As routine analysis and execution become automated, human value shifts toward judgement under ambiguity and governance.

A national literacy project is not just about the how-to. It is about the who-am-I. We need to educate the middle majority, the 60-70% of people currently caught between hype and horror stories, so they can move up the stack toward systems design and trust structures rather than protesting the inevitable tide.

3. Understanding the hierarchy of truth

The most dangerous hallucination is not an AI making up a fact. It is a human assuming a probabilistic model is a deterministic truth engine.

Real literacy in 2026 means understanding the difference between the Real World, broad cloud APIs, and My World, governed local context. We need to teach the public to demand a digital receipt: the metadata record that identifies the lineage of an answer.

Without this knowledge, we risk a mass hallucination event where truth becomes a relic.

The new national curriculum

If I were designing the BBC AI Literacy Project for today, it would focus on three pillars.

  • Sovereign parity: teaching people why bringing their primary compute home, through local LLMs and local-first infrastructure, matters for privacy and margins.
  • Agentic logic: understanding how to orchestrate multiple virtual coworkers rather than simply chatting with a bot.
  • Algorithmic governance: interrogating our reflexive biases and learning to manage systems where decisions carry real-world financial or legal consequences.

We are overestimating how easily AI can replace us, but drastically underestimating how it can amplify us. In the 80s, the BBC Micro was a tool to help us see the future. In 2026, AI is a mirror. It is time we learned how to look into it without blinking.

It is time.