Fawzooz
عربيCollaborate
Your journey
  1. 1
  2. 2
  3. 3
  4. 4
  5. 5
  6. 6
0 of 34 units

You are in Degree 1 · You are hereunit 1 of 4Ahead of you: Your own assessment matrix, coloured in by hand.

Degree 1 · Unit 1.1

A world that learns without you

In November 2022, millions of people opened a blank page, typed a question into it, and were answered as though by another human being. Most people took that day to be the birth of artificial intelligence. It was not. What was born that day was a simple interface to a technology that was already decades old. The intelligence was not new; your access to it was.

And because most people came in through that particular door, their picture of the whole field is still tied to a chat box: you ask a question, it gives you an answer. Four years later, that picture is no longer accurate. This is the difference that makes reading this today quite unlike reading any book published two years ago.

What actually changed: four shifts

Of the thousands of headlines that have passed in these years, only four of them genuinely change the way you work. Everything else is detail.

FIG. 1 — The timeline that concerns you
2022
Access — The technology came within reach of people who cannot program at all, so the barrier stopped being a technical one and became a barrier of understanding.
2023
Memory and size — The context window widened from a few pages to entire books, which meant you could give the system your whole file rather than just your question.
2024
Multimodality — Systems began to read images and hear sound, and with that same capability, convincing forgery came within reach of almost anyone.
2025
Reasoning — Systems that "think" step by step before they answer, so their analysis improved a great deal — and they went on being confidently wrong.
2026
Action — Agents: systems that do not answer you but act on your behalf. They send, they edit, they make bookings, they write code and then run it. And at that point liability changed completely.

From answer to action

When a system gives you the wrong answer, you still have a chance to question it, verify the information, or simply ignore it. But when a system takes the wrong action — sending an incorrect quotation to a client, deleting an important file, or approving a request without proper checks — the consequences may already be real before anyone notices. That is the fundamental difference between AI that advises and AI that acts. And that is precisely why the last three degrees of this programme matter so much.

Where AI actually works

AI applications appear when we get a computer to perform one of the four tasks of human intelligence — or some of them, or all of them.

FIG. B1 — The families of application, and what they do
The four categories of human intelligence
Perception
To see, to hear, to read
Inference
To understand what it perceived and draw the meaning out
Decision
To choose the most suitable path
Action
To carry it out in the real world
⟶
AI applications appear when we get a computer to perform any of these tasks, or some, or all of them — through code.
The more categories a system covers, the closer it comes to autonomy — and the more you answer for designing and supervising it.

The order climbs in autonomy: from "sees" to "infers" to "decides" to "acts". Each rung is more the system can do alone — and more that you answer for in designing and supervising it.

And what has not changed?

This is a question we rarely ask, though it may be the most important one. Three things remain fundamentally unchanged. AI cannot distinguish the truth from something that merely sounds convincing. It does not understand professional accountability — it has no idea what it means for your signature on a wrong report to end your career. And it does not fully understand your context: the history of your organisation, the nature of your clients, or the important conversations that were never minuted.

These are the things you bring to the table. Those who understand this use AI to amplify their own abilities. Those who do not risk letting it replace their own thinking — and in the process they lose twice over: their judgement, and the chance to get real use out of the technology.

Do this

  1. 1 — On paper. Write down three tasks you carry out weekly that take more than an hour. Against each one put: is it essentially knowledge, judgement, or relationship?

  2. 2 — On the tool. Ask an AI system to explain a subject you know thoroughly. Record: what it got right, what it got wrong, and what looked right and was wrong.

  3. 3 — In your field. Find a documented case where AI failed in your profession. Where was the point of failure: in the tool, or in whoever trusted it?