Fear of technology making your job obsolete: what the data says and what to do

We talk a lot about what artificial intelligence can do and less about how the people working with it feel. This guide brings together three recent news items that, read together, tell the same story: technology is already here, and what defines how each person and each company fares is what happens around it.

How many workers fear technology will make their job obsolete

Gallup asked 1,200 U.S. adults in its Work and Education survey (August 3 to 24, 2026, phone interviews; 507 of them were employed, with a margin of error of ±6 points). It published the results on September 15, 2026.

• 27% of workers fear technology will make their job obsolete.
• It is the highest figure since the question was first asked. It was previously 13% (2017), 15% (2018), 17% (2019), 15% (2021), 22% (2023) and 20% (2025). Gallup did not ask the question in 2020, 2022 or 2024.
• It worries those under 45 more: 34%, versus 19% among those over 45.
• It grew both among those with a college degree (25%) and those without one (29%).
• The traditional fears barely changed: that benefits will be reduced (33%), that they will be laid off (24%), that hours will be cut (21%) or that pay will go down (21%).

To read it calmly: it is a U.S. survey, it speaks of "technology" and not only artificial intelligence, and the worker sample is 507 people. It is not a measurement of Colombia, but it is a useful signal: the fear of being left behind already weighs more than several classic work fears. Read it at Gallup.

What could happen to wages in Latin America

On September 21, 2026, the IDB Group brought together heads of state and technology leaders, alongside the UN General Assembly, to discuss AI. In its press release it previewed figures from a report to be published in November 2026 (From Digitalization to Artificial Intelligence: Turning Promises into Productivity):

• Broad AI adoption with large productivity effects could raise the GDP of Latin America and the Caribbean by 5.1% in a decade. With limited adoption, only 0.3%.
• On wages, the outcome forks: they could rise between 2.3% and 5.3% if workers manage to move into expanding areas, or fall between 13.5% and 20.9% if they cannot.

The reading that interests us most is this: the result depends on something that is not technological, the ability to prepare and move talent to where value is created.

To read it calmly: these are scenarios ("could"), calculated by the IDB itself, and the full report has not yet been published, so the figures cannot be checked yet. Read the IDB press release.

What a company can do

This is what we propose at Método y Estrategia, based on what we see in the field. These are our own criteria, not recommendations from the studies cited.

1. Ask before assuming

Often what the team feels about technology is assumed. A short conversation, with three questions, already gives valuable information:

1. Which part of your work do you think technology could change in the next two years?
2. What would you like to learn to feel at ease with that change?
3. What do you need from the company to be able to do it?

2. Build commitment instead of demanding it

Tatiana Agudelo, from the Método y Estrategia team, tells how, when someone on a team came back to work after a medical leave, the first question was not about the certificate but "do you feel well enough to work from today?", and the answer could be no. Her idea: commitment cannot be demanded permanently, it is built. When people do not have to prove their worth all the time, they end up giving much more than is asked of them. Read it on LinkedIn (in Spanish).

3. Look at the whole job, not just the task

Jose Cardona, co-founder of Método y Estrategia, tells how he was sought out to develop software and the project ended up requiring much more than code: understanding requirements, designing an architecture, learning new services on the fly and, above all, knowing how to communicate. Because, in his words, "a misunderstood requirement costs more than any bug". When technology absorbs a task, the skills that gain weight are usually the ones surrounding that task. Read it on LinkedIn (in Spanish).

Two tool examples that show how much the "after" matters

AI interviews already available. Amazon Connect Talent, from AWS, became generally available on September 16, 2026, for now in two U.S. regions (Virginia and Oregon). Its AI agents run structured voice interviews at any hour and score candidates; recruiters receive scores, transcripts and evaluations. The question that remains open: which part of getting to know a person can be structured without losing exactly what is being sought? AWS announcement.

An AI agent in production needs someone to review it. On September 22, 2026, AWS introduced Amazon CloudWatch Omni, a tool to observe AI agents: it records every step an agent takes and comes with 17 built-in evaluators that score its answers on things like coherence or helpfulness. It is technical, but the underlying idea is not: automation does not end the day the agent starts working; someone has to be responsible for making sure it keeps doing it well. AWS announcement.