What does it actually mean to work alongside AI? Is it just prompting?

In partnership with

I’ve been pitching out my new AI literacy offering for the last 4 weeks now, and the first question I get asked is “is AI literacy just teaching prompting?”

Meanwhile, every business I speak to is drowning in new AI news, have a ton of tools they’ve handed out to their staff - and yet still can’t control the outputs. Some renegade employees are building their own bots whilst some couldn’t care less. And then there are some who see AI as a threat; “humans VS the machines” - and avoid it out of fear.

AI literacy isn't about learning to prompt better, though that does help (and no you don’t have to be Shakespeare to prompt well).

It's about fundamentally rewiring how people think about their work, and understanding what it takes to have a colleague who works in code but communicates in English. Spoiler alert: an AI colleague is not to be treated like a human one, nor a software tool, nor ‘Google’ - which are the current usage patterns.

Much like we’ve taught technology to speak English, we’ll need humans to step towards how technology thinks too. We don’t all need to learn code, but we need to learn systems thinking, chain-of-thought and coder-brain logic to some degree.

The patterns are similar:

Most companies have 4 types of AI users once leadership has supportively (or threateningly) mandated AI usage:

1. The Thinkers: they think with it, brainstorm with it and write with it - but it’s still more like chatting with your wildcard ‘ideas’ mate, rather than working with a productive co-worker.

2. The Renegades: they use AI to do everything, to the point of losing what made them great to begin with. They’re building shadow AI workflows and creating compliance nightmares in their excitement.

3. The Resisters: they avoid AI out of fear, stubbornness or thinking it’s too technical for them, and as a result they’re falling behind.

4. The Overwhelmed: they read AI news every day trying to keep up - but it’s a firehose, and they use ChatGPT like Google, getting mediocre results.

The shift in work as a result of co-working with AI:

We're not just adding new tools to old workflows. We're fundamentally changing how work gets done.

The old model: Human does everything, maybe with some software assistance

The new model: Human designs the system, AI executes the routine parts, human focuses on exceptions and strategy

Why your current training doesn’t work

Most AI training treats people like they're learning Excel 2.0. But AI isn't a spreadsheet (it’s not even good at them) - it's a thinking partner that needs to be directed - not just operated.

The gap:

• We teach prompting techniques but not prompt strategy

• We show features but not workflow integration

• We focus on outputs but ignore the thinking process

• We train individuals but not teams on collaboration with AI

It's like teaching someone to use a phone, but never explaining how conversations work.

The business cost of getting this wrong

Here's what I see happening when companies skip the mindset shift:

• The $100K wasted on an annual ChatGPT Enterprise subscription that saves just 2 hours per week because people use it like an expensive search engine.

• The "AI-resistant" middle layer of management whose junior team quietly works around them, creating knowledge silos and eventually a stripping of the middle layer

• The compliance incident where someone fed sensitive data into a public AI tool because nobody taught them to think about data flow.

What Real AI Literacy Looks Like

True AI literacy means developing what I call "hybrid thinking" - part human intuition, part machine logic: