The audio controls above still work. Interactive translation, saving and practice need JavaScript. You can read the complete transcript and download your study materials below.
Complete English transcript
It is Monday morning.
You arrive at work and find a new AI tool on your screen.
By Friday, it can do part of the work you usually do.
Should you feel excited, worried, or both?
Both would be a very reasonable answer.
And what should you actually do next?
Welcome to Your English Toolbox.
I'm Martin.
And I'm Julia.
Today we are talking about AI and the job market.
Not just what might happen one day.
We have real examples of companies changing how people work.
And we have practical steps you can try, whatever your job.
Theo is here with us.
Hello, Martin and Julia.
I'll explain the examples, and I'll also share my opinions.
When something is my opinion, I'll say so.
Let's start with someone answering questions from customers.
We will call her Sofia.
Sofia works on a customer support team.
Every day, people ask when a delivery will arrive or how to return an item.
Then the company introduces a chatbot.
It begins to answer some of the simple questions.
If I were Sofia, my first question would be: am I still needed?
Of course.
That is not a theoretical question when it is your rent and your future.
Has something like this happened in a real company?
Yes.
Ingka Group is the largest IKEA retailer.
It introduced a chatbot called Billie for customer enquiries.
An enquiry is a question from a customer.
So what did Billie do?
The company said that, between 2021 and 2023, Billie resolved around forty-seven percent of the questions sent to the chatbot.
Wait.
Forty-seven percent of every customer question at IKEA?
No.
That figure refers to the questions Billie received.
It is an important difference.
And what happened to the people?
Ingka said it had trained eight thousand five hundred customer-support colleagues in other skills.
Those skills included remote interior design, digital sales and handling more complex questions.
So imagine Sofia no longer spends so much time answering the same simple questions.
She could instead help a customer plan a room or solve a problem.
That sounds much more interesting.
But does the report prove that no one lost a job?
No, it does not.
We should not turn one company announcement into a promise about every worker.
And another company could make a different decision.
Exactly.
A tool does not choose the company's plan.
People do.
Let's make that practical.
Suppose my manager says a chatbot will join our team.
What could I ask?
You could ask: “Which tasks will change?”
Not just: “Will my job change?”
Yes.
A job contains many tasks.
A tool may change some of them before it changes the whole role.
I would also ask: “What training will we get?”
That is a good question.
And ask when the training will happen.
Why is the time important?
Because not everyone has free evenings or money for a course.
A company can give people time and support during the working day.
Let's practise the question together.
“What training will we get?”
There is another difficult question for Sofia.
What if the new work looks better, but she has never done it before?
She may be excellent with customers and still feel nervous about a new tool.
Of course.
Learning a new task takes time, and people need a chance to practise.
Let's imagine her manager offers a short course in online sales.
Sofia wants to know what happens after the course.
She could ask: “Will I practise with a colleague?”
And: “Who can help if I get stuck?”
That makes the training more concrete.
It is not only a presentation followed by a new target.
Exactly.
In my opinion, good training includes practice, feedback and time to ask questions.
Could Sofia use the skills she already has?
Certainly.
She already knows how to listen to customers and explain a problem.
Those skills can support a new role, even if she needs to learn different tools.
So adapting is not necessarily starting again from zero.
That is a useful way to think about it.
But let us be honest.
Training does not guarantee that every person will keep the same hours or salary.
That is right.
People also need clear answers about their contracts and responsibilities.
Sofia could ask: “How do you expect my role to change?”
A question like that helps her understand the actual offer.
We have looked at customer service.
What about someone who has an idea but cannot build an app?
That sounds like me.
I have lots of ideas and absolutely no idea how to code.
Then let's look at Duolingo.
The company says its product managers used AI to help make a prototype for Duolingo Chess.
A prototype is an early version you can test.
So one person had an idea and AI built the entire final app?
That is not what the company says.
Its product managers used AI to build a prototype.
The engineers then built the final product.
So the prototype helped people show an idea sooner.
Yes.
They could look at something concrete and discuss what needed to change.
What does that mean for the people on the team?
In my opinion, more people may be able to test an idea before handing it to a specialist.
But a working prototype is not the same as a safe, reliable final product.
And someone still has to decide whether the idea solves a real problem.
I can imagine a designer asking: “Who will use this?”
Or an engineer asking: “What happens when it fails?”
Exactly.
The skills can overlap, but that does not mean every specialist becomes unnecessary.
What about someone starting their first job?
Junior workers often learn by doing the simpler tasks.
That is a serious concern.
If a tool does the first draft, a beginner still needs a way to learn how the work is done.
For example, a junior colleague could review a draft with someone more experienced.
And ask why one answer is useful while another is wrong.
Yes.
In my opinion, a team should make space for that learning rather than assume beginners can skip it.
We have talked about support teams and apps.
Does AI only change office work?
No.
Consider technicians who visit homes to repair appliances.
A washing machine breaks, and someone comes to fix it.
Exactly.
Haceb is an appliance company in Colombia.
Microsoft describes an AI assistant called Wilson that helps its technical teams find information about products and problems.
So a technician can ask for help while working with a customer?
That is the idea described in the case.
The tool helps find information, but the technician still works at the customer's home.
Could the assistant suggest the wrong solution?
It could.
A person still needs to check the model, assess the problem and use their professional judgement.
That sounds important in many jobs.
A fast answer is useful only if it is reliable.
And a customer may need a person to explain the next step.
Yes.
Technical knowledge, judgement and communication can all matter together.
Now let's talk about our listener.
Maybe you do not work in customer service, design or repairs.
Maybe you are a teacher, a driver, a nurse or a student.
Where do you begin?
Start with one task you understand well.
Write down what you do, what information you use and how you know the result is good.
For example, preparing a simple weekly report.
Yes, if that is part of your work.
Then check your workplace rules before using any tool.
That includes rules about customer information.
Exactly.
Do not put private information into a tool unless your organisation has authorised that use.
Could I practise with an imaginary example instead?
Yes, that can be a useful place to start.
Try a small example that contains no confidential information.
Then compare the result with what you would normally produce.
Not just how quickly the tool answered.
Right.
Check accuracy, missing details and the time needed to correct it.
What if I do not know which tool my employer allows?
Ask.
You could say: “Which tools are approved for this task?”
And if the answer is none?
Then follow that rule.
You can still discuss what training or support would help your team.
Let's give our listeners another useful question.
“How do we check the result?”
Theo, some people say everyone must use AI for everything.
What do you think?
I think curiosity is valuable, but I would not promise that every tool improves every task.
The best choice depends on the work, the risk and the people affected.
So trying a tool and trusting it blindly are different things.
Very different.
A tool can help you explore an idea and still need careful checking.
Let's return to Sofia.
Her team has a new chatbot, and she is wondering about her future.
She could ask which tasks will change.
She could ask for training time and support.
And she could ask: “What happens to the people doing this work now?”
That is a question for the company, not something Sofia can solve alone.
Exactly.
Workers can learn, but employers also make choices about training and opportunity.
If AI saves a team time, that time could be used in different ways.
A business might reduce costs, grow, or offer better service.
And these choices can have very different effects on people.
In my opinion, companies should explain those choices honestly and give people a chance to adapt.
What is your first step this week?
Choose one task you know, check the rules, and think about the results you need.
And keep one question ready for a real conversation at work.
“What training will we get?”
Thank you for listening to Your English Toolbox.
Thank you for having me.
Take care, and keep learning.