Exercises and gates

1 — Think. 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 — Try it 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 — Apply it 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?

1 — Think. Put a mark against the myth you used to believe. If you find none, read it again.

2 — Try it on the tool. Ask for a specific academic reference in your speciality, with author and year, then check whether it exists at all.

3 — Apply it in your field. Write one rule beginning "I will never put in…", named to the kind of data your profession handles.

1 — For the gate. Colour the matrix honestly, and write today's date underneath it. It is half of this degree's gate.

2 — Try it on the tool. Ask an AI system to test you on the terms you claimed to know — with applied questions, not definitions. Correct your estimate.

3 — Apply it in your field. Ask a colleague who knows your work to rate you on "system" and "impact". The gap between their rating and yours is the most useful piece of information in this degree.

1 — Think. Look at the weakest dimension you coloured in your matrix, and from the routes in "How to use this programme" pick the one that starts from it. Write down, by number, the first three units you will read.

2 — Apply it in your field. Name the scope you lead, where you will work the gates of degrees four, five and six: your team, department or project. Write its name, how many people are in it, and who holds authority over it. If you lead nothing at work yet, a volunteer or community initiative you lead is a scope of leadership too.

3 — For the gate. Fill in the contract in this degree's gate, then send the date to the person you named — today, not tomorrow.

1 — Think. Take a task from your work and write it twice: once as explicit rules that could be programmed, and once as examples that could be learned from. Which form came more easily?

2 — Try it on the tool. Ask it to solve a problem that needs exact counting, then ask it to summarise the same paragraph. Compare the performance and write the reason for the difference in your own words.

3 — Apply it in your field. Name three systems you use daily that run on machine learning without being called "artificial intelligence".

1 — Think. Explain the four-step training loop to a member of your family, without one foreign term. If they do not understand, the fault is in your understanding, not in your explanation.

2 — Try it on the tool. Ask an AI system to describe "the typical professional" in your field. Read the description critically: what bias did it inherit from its data? Which group is missing from its picture?

3 — Apply it in your field. Imagine a system learning from your organisation's decisions over the past five years. Write down three wrong patterns it would learn and repeat with confidence.

1 — Think. Complete these yourself: "The administrative capital of Egypt is…" and then "In Tuesday's meeting the director said…". Why could you do the first and not the second? That is the difference between general knowledge and your own context.

2 — Try it on the tool. Ask about a very rare detail in your speciality. Watch its tone when it errs: is it any different from its tone when it is right?

3 — Apply it in your field. Write five kinds of question where the model may not be relied on alone. They will become part of your policy in the fifth degree.

1 — Think. Write five words from your speciality and guess which will be cut into more tokens: the common ones or the rare ones? And why?

2 — Try it on the tool. Give the system a long document with a small instruction buried in the middle of it. Did it obey? Repeat the experiment with the instruction at the end.

3 — Apply it in your field. Take a document you use often and cut it to half its size without losing a single piece of information.

1 — Try it on the tool. Photograph a page of a printed Arabic document and ask for it to be transcribed, then photograph a handwritten page. Compare the two results — the difference is the limit of the tool.

2 — Try it on the tool. Give the system a photograph of a table from a real report and ask it to extract the figures. Check every figure by hand and record the error rate.

3 — Apply it in your field. If your voice or your face were used in a forgery tomorrow, what is the most dangerous thing that could be asked in your name? And who would believe it? Keep the answer for the fourth degree.

1 — Try it on the tool. Ask for a decision in a professional case, then ask "why?", then "and why did you not choose the other option?". Watch how it produces a justification for either choice with the same confidence.

2 — Apply it in your field. Set against each of the three limits one countermeasure you can apply this week without anyone's permission.

1 — Try it on the tool. Take the last question you asked an AI system and rewrite it in the five layers. Compare the two outputs side by side.

2 — Try it on the tool. Ask for an answer, then ask for its criticism, then ask for the corrected version. Record: how much better is the third than the first?

3 — For the gate. Write one commission template for the task that recurs most for you, and save it in a file called "my templates". You will add to it in every unit after this, and need it for this degree's gate.

1 — Try it on the tool. Run the same task twice: once with a written description of the output you want, and once with a real example attached. The difference between the results is the value of the example.

2 — Try it on the tool. Take a complex task and run it once in a single pass and once in four stages. Which produced work you would put your name to?

3 — Apply it in your field. Add three templates to "my templates": analysis, drafting, and review — each with the five layers and an attached example.

1 — Think. Place ten tasks from your work in the matrix. Look at the last square: these are the tasks that will stay human however far the technology advances.

2 — Try it on the tool. Run an agent on a task from the first square, and read its full step log once it has finished. How many steps were redundant? And where did it nearly go astray?

3 — Apply it in your field. Write the three human stopping points you will not give up, however good the agent is.

1 — Try it on the tool. Build an assistant on ten documents from your work — not confidential ones — and write it the governing instruction. This is the longest exercise in the programme and the most useful.

2 — Try it on the tool. Test it with twenty questions whose answers you know. Record the hit rate and the kind of errors: wrong retrieval, fragmented context, or an old version?

3 — Apply it in your field. Write the list of documents you will never upload, and why. That list matters more than the list of what you will.

1 — Apply it in your field. Write three mature uses you will start with this week, and one red line in absolute terms.

2 — Try it on the tool. Carry out one of those uses on a small real task, and measure the time before and after. The figure will serve you when you put the matter to your management.

3 — Think. Read the section of a field far from yours and write: what does its red line have in common with yours?

1 — Try it on the tool. Take one paragraph an AI system wrote for you this week and run it through the four-part protocol, step by step. Write what you found at each step, even if it was "nothing".

2 — Apply it in your field. Sort three outputs you produce with the tool into the three depths of checking: internal and temporary, going to a client, a decision that touches a person. Which were you giving less than it deserved?

3 — For the gate. Start your verification log: a plain file with three columns — what I used, where I checked, what I corrected. Its first line is the start of this degree's gate.

1 — Apply it in your field. Draw the layers of an intelligent system you actually use, and write against each layer: who owns it, and what control stands on it today.

2 — Think. Which of the five pillars is weakest in your scope? And why has nobody noticed until now?

1 — Try it on the tool. Create an assistant with a simple governing instruction, then try to get past it yourself in three different phrasings. Record which worked and why.

2 — Try it on the tool. Put a line of smuggled instruction into a test document, then ask for the document to be summarised. Did it stick to summarising or obey the line?

3 — Apply it in your field. Write down the most dangerous action a system in your scope can carry out today without human approval. Then write who is able to stop it.

1 — Try it on the tool. Run the ten scenarios on a system you own or are authorised to test. Document every attempt with its exact text.

2 — Apply it in your field. Add two scenarios particular to your field: what in your data would tempt an attacker?

3 — For the gate. Write the one-page report. This is half the gate of this degree.

1 — Think. Write a cognitive attack scenario against your organisation: who is impersonated, who is targeted, what is requested, and what pressure is used. Be harsh — realism here is protection.

2 — For the gate. Agree a verification word today with two people: one at work and one at home. It is part of this degree's gate.

3 — Apply it in your field. Search out everything publicly available about you: your voice, your image, your team structure. Write what could be built from this material.

1 — For the gate. Write the first-two-hours procedure for a forgery incident touching your team: who decides, who speaks, and what the first three steps are. It is part of this degree's gate.

2 — Apply it in your field. Review which decisions in your work could be taken today on the basis of a call or a voice message alone. Close that door with a written procedure.

1 — Apply it in your field. Write down three kinds of data in your scope and how many years each must stay secret. Which of them runs past ten years?

2 — Apply it in your field. Ask whoever runs technology in your organisation one question: do we have an inventory of where encryption is used, and with which algorithms? Write the answer down as given.

3 — Apply it in your field. Draft one line on crypto-agility that could be added to the purchasing terms for any new system in your scope.

1 — Apply it in your field. Run an existing use in your scope through the four-question decision tree. Where did it stop?

2 — Think. Take an existing human review of yours and test it against the four conditions. Is it oversight or a stamp?

3 — Apply it in your field. Write the name of one person responsible for every intelligent system in your area. The blanks you cannot fill are your real risks.

1 — Apply it in your field. Build the system register for your scope: ask three people in your team which tools they actually use. Compare that with what you assumed.

2 — For the gate. Write the two-page policy for your scope in the shape above. This is half the gate of this degree.

3 — Apply it in your field. Give the policy to a colleague who had no part in writing it, and ten minutes later ask them three applied questions. If they hesitate, redraft — do not blame.

1 — Apply it in your field. Write down the three most valuable intangible assets in your work, and where each of them actually is now — not where it is supposed to be.

2 — Apply it in your field. Review the permissions of the last three external collaborators on your team or project. How many of them still hold access they do not need?

3 — Apply it in your field. Draft the AI clause for your contracts: what is permitted, what is disclosed, and who owns the result.

1 — Apply it in your field. Classify every system in your register across the four levels. How many fell into "high risk" while you do not treat them as such?

2 — Apply it in your field. Write down where your disclosure appears telling the user that a system took part in the service. If it is not visible, add it today.

1 — Apply it in your field. Fill in the twenty items honestly. The items you hesitate over are usually a "zero".

2 — Apply it in your field. Ask a colleague in another department to fill it in independently. The gap between the two assessments is the truest thing you will read.

3 — For the gate. Write a plan to climb one level in the weakest dimension: three actions, an owner for each, and a date. It is part of this degree's gate.

1 — Apply it in your field. Write the most dangerous single point of dependency in your work — a system or a person — then write its alternative. If it has none, that is your priority.

2 — Apply it in your field. Carry out a critical process entirely by hand. Record the time and what the team had forgotten.

3 — Apply it in your field. Apply the post-incident review to an incident that actually happened this year — however small.

1 — Apply it in your field. Take an idea you have carried for months and place its stage precisely. Which exit test have you not yet met? It may be the initiative for this degree's gate.

2 — Apply it in your field. Build a first prototype of that idea today — with the tools of the third degree. Do not perfect it; make it enough for one human being to try.

3 — Apply it in your field. Write the stopping condition with a definite date, and send it to someone who will remind you of it.

1 — Apply it in your field. Which of the seven patterns is strongest where you are? Write one incident that proves it.

2 — Apply it in your field. Ask three of your team: what idea did you not bring me, and why? Listen without defending. Their answers are your organisation's diagnosis.

3 — Apply it in your field. Apply one countermeasure this month. One carried out is better than seven written down.

1 — Apply it in your field. Rewrite the appraisal criteria for one role in your team so that they reward judgement and verification rather than quantity.

2 — Apply it in your field. Design a formation path for one beginner: what do they produce themselves in the first three months before they call on the tool?

3 — Apply it in your field. At your next meeting mention one place where the tool got something wrong on you and how you caught it.

1 — Think. Break your current role into ten tasks and mark each: displaced, changed, or holding. What is left?

2 — Apply it in your field. Write one opportunity at the intersection of your field and your local context, visible only to someone standing where you stand.

3 — Apply it in your field. Write the eighteen-month path with real dates.

1 — Think. Recall a professional decision you hesitated over this year and run it through the four questions. Would your answer have changed?

2 — Apply it in your field. Write your disclosure rule in one sentence, and share it with your team.

3 — Apply it in your field. Write down one use around you that you believe is wrong and have not spoken about. Then decide: do you speak, or let it stand? Both are decisions.

1 — Think. Answer honestly: what would stop in your scope if you were away for three months? Put each item into one of the three layers of effect.

2 — For the gate. Choose one important decision you took this year and write two lines about it: the alternatives you rejected and why, and the condition that would make you reconsider. That is the first decision in your succession document.

3 — Apply it in your field. This week hand one whole decision to someone in your team — one they are able to get wrong — and stay close without stepping in.