No need for AI at work? Start with what you already try for yourself

Many people enjoy using AI, just not for work. They discuss what they have read, get to grips with an unfamiliar topic, come up with something for themselves. And when asked about work, they say: "Everything is set up there. I don't see why I'd need it."

There is no need to argue with that. At Simplit, we start with a different question: is there a task at work that you genuinely want to handle differently?

To find out, start with a familiar personal task. Pick out the action that turned out to be useful and look for a counterpart at work.

Personal interest does not need a business case

A conversation about AI easily acquires an obligatory sequel: how to make money from it, what to automate, how much time to save. But you can open the app for other reasons too: you are curious, you want to understand a difficult text, you enjoy coming up with options and talking them through.

If AI helped you get to grips with a topic you had long been putting off, that is a good enough result. You do not have to turn it into a project, a system or a source of income. Otherwise every hobby starts to look like raw material for productivity: the person has not yet had a chance to enjoy the activity, and they are already being offered a process with metrics. You can start with what interests you and ask the question about work later — if there is a reason to.

With a familiar task, it is easier to see what happened

When AI produces an impressive result in an unfamiliar field, it is hard to judge: the answer sounds convincing, but you do not know what is missing from it. With a familiar task, you have more to hold on to — you know why you took it on, what matters to you and what result would not work for you.

An example. You are planning a short trip and have saved descriptions of several routes. One is more interesting, another fits your schedule better, and the third passes through a place you have long wanted to visit.

You could ask AI to pick "the best route", but the word "best" hides almost everything the decision depends on. One person wants to see as many places as possible, another needs a relaxed pace, a third is ready to drop half the programme for the sake of a single event.

It is more useful to state your conditions first and ask for the options to be compared against them: showing the transfers, free time and the information still missing for the choice separately.

After the answer, you can spot specific mistakes. For example, AI called a route relaxed even though you would have to move every day. Or it took the lack of information about a road to mean that the road does not exist. You understand why this matters and adjust the request — and that shows what helps you get a usable result.

Timetables, availability and other changing conditions still need to be checked against current sources: a neatly assembled plan does not confirm them.

The same action can sit behind different topics

In the trip example, you did several quite specific things: you gathered scattered information, named your criteria, asked for the options to be compared, checked whether an important condition had been missed, and separated what is known from what still has to be found out.

Remove the word "trip" from the description. What remains is a task: compare several options against my conditions and show the gaps that stand in the way of a decision. That is already easier to compare with work.

Say you are reviewing proposals from several contractors. One describes the scope of work in detail, another the deadlines, and in the third some of the terms are missing altogether. Before discussing the choice, the information has to be brought into a comparable form and questions have to be drawn up. The topics differ, but the action is similar: build a comparison against set criteria and do not present the unknown as known.

This does not yet mean that AI can choose a contractor or that you can hand all the documents over to it. All you have found is an analogy worth testing. That is exactly the level at which to look for a use at work: name the action that turned out to be useful and see whether it occurs in another part of your life.

When carrying it over, work through the conditions of the new task

After a successful personal attempt, it is tempting to save the request that worked and reuse it for other tasks. Sometimes such a template really does help you get started, but you will still have to work through the conditions of the new task.

On a trip, you decide for yourself. In a work decision, other people may be involved with their own requirements, some of the information must not be passed to an external service, and some terms only make sense together with additional documents or agreements.

Even a familiar comparison table can let you down. One proposal says "support included". If AI puts a confident "yes" against that item, the table will look tidy, but the comparison will be misleading. The useful result here is a question: what exactly does the support include, for how long and where is that written down?

The principle from the personal task applies here too: keep missing details as a gap. But the comparison criteria and the way you check will be different. A similar structure helps you choose what to try, but it does not prove that the tasks are the same.

Test the work hypothesis separately

Suppose you have found an analogy: in the personal task it was handy to compare options, and at work this need comes up regularly. Limit the first attempt to a draft comparison based on a small set of materials that may be used in the chosen service.

Define the result in advance — for example, a table with the same criteria, the source of each term and a separate list of questions. The decision on the choice stays with a person.

Then check it against the source documents:

• whether all the essential terms made it into the comparison;
• whether a basis for each item can be found in the source;
• whether an assumption has turned into a fact;
• whether it has become easier to prepare for the discussion;
• how much effort went into preparing and checking.

The answers vary. The draft may genuinely help you see the differences. It may turn out that the main work still stayed with you. Or the documents may be so different that you first need to agree on the criteria themselves. The outcome of the check is a decision to continue, change the instruction or keep the usual way. The personal attempt gave you a reason to try, but its usefulness at work has to be established anew.

When no analogy turns up

Sometimes a person enjoys using AI outside work but sees no suitable work task. Do not immediately put this down to a lack of imagination. Perhaps the work is organised differently, the necessary data is not available, preparing and checking takes more effort than the usual way, or there is simply no task right now that is worth changing anything for.

A list of possible uses can be made up for almost any profession, but an item on such a list is not yet a reason to spend time on it. You can leave the question open. When a familiar difficulty comes up, you will come back to it with personal experience: you will already know where AI helped and what not to expect from it without checking.

Exercise: describe the action without naming the topic

Take one personal task that you have already done with AI or plan to try — one where you understand the result and would notice a significant mistake. Answer five questions.

1. What did I want to get? Name the result in plain words: compare options, understand an explanation, organise notes, prepare questions.
2. Which part did AI do? Separate it from your own work. Perhaps you found the sources, set the criteria and made the decision yourself, while AI put together the first version of the comparison. That is already specific enough.
3. What did I have to explain or correct? These are the conditions without which the approach does not work: which information counts as confirmed, what must not be made up, why one criterion matters more than another.
4. Does a similar action occur at work? Look for similarity in actions — comparing, grouping, explaining, finding gaps, preparing a draft — not in the name of the topic or document.
5. What will change when it is carried over? Who will use the result, which materials can be shared, who will check the answer and what happens if a mistake goes unnoticed.

If the answers point to a small, testable task, formulate a hypothesis. For example: "In my personal task, comparing options against my criteria helped. At work, I will try preparing a draft comparison from the permitted materials in the same way, check it against the sources and see whether it has become easier to put together questions for the discussion."

If there is no similar task, you can stop there — you have still understood your own experience more precisely.

How to start the conversation

You do not have to come to a conversation about AI with a request to "bring AI into the business". It is enough to say: "I tried this for myself, it was interesting, but I don't see a use at work yet." From there, we can work through what you did, where the benefit lay and what conditions made it possible, and then see whether there is a similar task at work and what is different about it.

The phrase "I don't need AI for work" may remain your conclusion. Or it may become more precise: "I'm not going to change all my work, but I want to try this one area." That kind of interest is enough for a useful conversation.

Read next: if you would rather look for your first task not in personal experience but in your working week — "Already paying for AI? Choose one recurring task for it". In both cases the check is the same: what came out of it, how it helped and what was left for a person to do. How to describe a task in your own words and check whether AI has understood it correctly — "How to turn a description of a problem into a testable task for AI".

You can work through your own example with Simplit in a free 30-minute meeting: tell us what you tried for yourself and where you see a similarity at work. Discuss your task