You will be able to score a task on checkability, cost of error and fit with what models do well.
Marcus is an account manager at a freight forwarding company in Changi. On a typical Monday he writes a status update to each of his six clients, prepares a short volume report for his manager, answers a complaint about a delayed shipment, and drafts notes for a quarterly review. After lesson 8.1 he knows AI will do some of these well and fumble others, and that the difference will not be obvious. What he needs is a quick way to tell which is which before he finds out the hard way.
This lesson gives you three questions that sort any task. None needs technical knowledge. Each is scored from one to five, and in every case a higher score means the task is better suited to AI.
Start with checking, because everything else depends on it. Module 6 showed that any AI output can contain errors that look exactly like correct content. A task suits AI when you can check the result faster than you could produce it.
A draft email to a client you know well scores high. You can read it in a minute and see at once whether it says what you meant. A summary of a meeting you attended is similar, because you know what was said. Those might score a four or five.
An analysis you would have to redo to verify scores low. If Marcus asks an assistant to work out which of his clients are shipping less this quarter, the only way to know if it is right is to go through the figures himself. At that point, the AI has saved him little and added a risk. That kind of task might score a one or two.
Next, picture the output going wrong and ask what happens. An awkward sentence in an internal note costs almost nothing. Someone frowns, and you fix it. That scores a five.
A wrong figure in a client report costs more. It can damage trust, lead to a bad decision, or take a week of emails to correct. A wrong statement in a legal letter, a regulatory filing, a price quote or anything involving someone's money or health costs more again. Those score a one or two.
Cost depends on where the output goes, not just on the task. A rough summary of a contract for your own understanding is low cost. The same summary pasted into an email to a client becomes high cost, because now someone else will rely on it.
The third question draws on the whole course. Language models are strong at work built from language: drafting, summarising text you provide, rewriting in a different tone, reformatting, translating between plain and formal writing, and brainstorming options. Those tasks match what lessons 3.2 and 4.1 described, where the likeliest text is usually useful text.
They are weaker at fresh facts unless they can search (lesson 4.4), exact figures and references (module 6), long documents where one buried detail matters (module 5), and judgement about people, such as whether a client is about to leave or whether a candidate suits the team. Tasks of that kind score low on fit.
For Marcus, the client status updates score five on fit. The volume report scores two, because it depends on exact numbers. The complaint reply is somewhere in between: the wording is a good fit, but the facts about the delay must come from his own records.
Some tasks score well on all three questions and still belong to you. The scores measure accuracy and effort. A few things matter beyond those.
Accountability: if your name, signature or professional licence sits on the output, you need to be able to stand behind every line, and saying the AI wrote it will not help you. Confidentiality: client data, staff records or anything covered by an agreement or by the PDPA may not be yours to paste into an outside tool. Check your company's policy before you do. Relationships: a message of condolence, a difficult conversation with a staff member or a personal apology needs your own words, because the person receiving it will be able to tell.
Marcus decided the complaint reply was one of these. The facts were simple enough, but the client had been with the company for years, and he wanted the apology to sound like him.
Put the three scores side by side and you get a quick ranking. Tasks scoring high on all three are your best candidates to hand over, and lesson 8.3 turns that into three clear zones. Before then, try the scoring on five tasks from your own week. It goes faster than you would expect, and the ranking is often surprising.
Score five of your tasks from one to five on checkability, cost of error and fit, and see which ones rise to the top.
Junxiong-WFG Organisation is an authorised representative of AIA Financial Advisers Private Limited (Reg. No. 201715016G).