Mar 22, 2026
3 min read
A report from Anthropic shows that computer intelligence will not replace manual jobs such as engineers or farmers.What professions are at risk?
Artificial Intelligence Can't Be a Mechanic or a Farmer: An Employment Myth Debunked by Anthropic
A report by Anthropic reveals that artificial intelligence will not replace manual jobs such as mechanics or farmers.What professions are at risk?
Arguably the biggest unknown about artificial intelligence is that there are those who wonder whether it will take our jobs, dominate the world and even destroy people.Since the second one looks more like an episode of a futuristic series, we'll talk about the first one for now.
Anthropic recently released a report examining AI's ability to disrupt employment.This text is just one of hundreds being published to examine whether or not AI is stealing jobs across industries.But it is worth dwelling on its content.
Why is it different?Anthropic wanted to demonstrate once again that AI is still far from filling our jobs, and that real-world uses of AI may not reach its theoretical capability.
The predictions of Dario Amodei: an investment project?
Since he left OpenAI for ethical reasons, Dario Amodei has always vindicated the danger that a technology like ChatGPT's generative language model can be uncontrolled.This is where Anthropic was born and then Claude, a leader of true AI professionals, especially for the IT sector.
Amodei warned at the World Economic Forum that "AI models will be able to handle software development" within 6 to 12 months. He also assured in 2024 that AI could be "as good as a Nobel Prize winner" in 2025, 2026 or 2027. So, is Amodei contradicting himself with these ideas about the development of artificial intelligence?
Bruno Marcial, Director of AI and Professor at EBIS Business Techschool, explains his opinion on the matter for Business Insider Spain.The professor emphasizes that “the business model of these companies is based on the consumption of tokens, so the more usage is encouraged, the higher their turnover will be.”In fact, he claims that “trends launched by companies like ChatGPT, OpenAI or Grok themselves, lead to competition to create the best model of the most attractive images or videos.”
In this case Amodei's statements could concern his IPO.
“Billion-dollar investments in AI will not only go to data centers, but will build a compelling story that justifies the data centers themselves,” said Jonathan Ammenidou, CEO and founder of WelPay.
In fact, referring to large AI companies, "creating FOMO (fear of missing out in English) among managers and boards of directors is the most effective way to accelerate the adoption curve when you're competing to be on the ground before your opponent," Amenedo asserts.
The highest paying jobs are the most at risk
According to the chart, among the highest-risk occupations in Anthropic's observed exposure index are computer programmers, which account for 75% of roles.They were followed by customer service representatives, data entry operators and financial analysts.
But the most remarkable thing is the opposite extreme.What the Anthropograph shows is that the jobs least at risk are cooks, mechanics, lifeguards, waiters, dishwashers and locker rooms.The ones that AI can't cover even a little at the moment are motorcycle mechanics and farmers.
Amenedo emphasizes that the real danger is not "a lack of work", but "the mentality of professionals who give their standards to models without knowing that this step is what really makes them valuable."In addition, he emphasized that at this time "responsible parties have stopped developing the ability to detect errors."
In this context, according to data from the current US population survey, AI threatens skilled and high-paid workers more than low-income workers.
Is AI destroying jobs now?
At the moment, this technology is not working systematically.Marcial pointed out that "the fear is constant, but we must understand that the companies that use this technology will be more efficient than their competitors, and those that will not work slowly."
A study by Anthropic found no significant increase in unemployment among the most exposed workers since the introduction of ChatGPT in late 2022. However, the most relevant search points for workers between the ages of 22 and 25, who are finding it increasingly difficult to break in, are.
"Today, it will not eliminate our work, but in two, three or five years, there will be a demand for change in different areas and services," Marcial emphasizes.
"AI is technology, and technology itself does not create evil: the creator of evil is the person who uses it. Therefore, we should not be against technology, but against its use," he concludes.
