Aleksei Mogilnikov
Back to blog
•
3 min read
What the Data Says About AI and Jobs

A developer working as a wave of AI noise closes in

There is a lot of noise around AI:

Anthropic/OpenAI just killed <a multibillion-dollar company>

(c) a LinkedIn post

<profession> will no longer be needed after 2024/2025/2026/2027

(c) CEO of an AI company

We have already entered will soon enter the singularity

(c) CEO of another AI company

It feels like technology and markets are changing radically, and everyone around you is already living in the future. Meanwhile, you are still digging a hole with a stick and hanging a bronze pot over a fire. That is where FOMO and anxiety come from.

What does the data say?

Is everyone using AI?

  1. Only 15% of workers use AI daily [1].
  2. Only 19.8% of companies used AI in at least one business function [2].
  3. Even in tech, only 42% of workers use AI daily; in most other sectors, the figure ranges from 9% to 15% [1].

AI is already being used, but it has yet to become an everyday tool for most companies and workers.

Is AI replacing people?

  1. Only 34% of employees at companies that have adopted AI agree that it has changed how work is organized at their company [1].
  2. Among companies where AI affected employees’ tasks, 66% used it only to augment employees’ work [3].
  3. Among all companies using AI, only 10% reported that it performs tasks previously done by employees [3].

AI more often augments employees’ work than takes over their tasks.

Is AI putting people out of work?

  1. Only 2% of companies using AI reported a reduction in headcount [3].
  2. No significant changes in employment or wages were found in occupations exposed to AI [4].
  3. In Denmark, no significant effect of AI on average earnings or working hours was found in the first two years after ChatGPT launched. Effects greater than 2% were statistically ruled out [5].

The data does not show an overall decline in employment due to AI.

AI is changing work, but not as quickly or as widely as LinkedIn feeds and the media make it seem. You can take your time learning the tools you need without trying to keep up with every prediction.

[1] https://www.gallup.com/699797/indicator-artificial-intelligence.aspx
[2] https://www.census.gov/library/stories/2026/05/ai-use-businesses.html
[3] https://www.census.gov/library/working-papers/2026/adrm/CES-WP-26-25.html
[4] https://budgetlab.yale.edu/research/what-we-do-and-dont-know-about-how-ai-affecting-labor-market
[5] https://www.nber.org/papers/w33777