Humanity CEO Dario Amody is worried that AI will be able to eliminate half of its entry-level white-collar work in five years.
There are many other potential labor risks lurking, including the threat of massive tariffs on imports in the US, layoffs aimed at revoking Australian construction during the pandemic era, and policies that empower employers at the expense of the workforce.
However, Amodei appears to believe policymakers are “sugar-inducing” the possibility of major AI-driven employment disruptions. This is a polite interpretation of the Trump administration's unquestioned AI policy, which threw Biden-era AI safety rules through an executive order, celebrating AI infrastructure deals.
Amodei expressed this concern in an interview with Axios last week. In code with Claude, the first developer conference of humanity. He suggested that unemployment could reach 10-20% over the next 1-5 years.
“As producers of this technology, we have a duty and an obligation to be honest about what's coming,” Amodei told Axios. “I don't think this is on people's radar.”
In fact, this concern appears in a variety of reports, such as the World Economic Forum's Future of Jobs Report 2025. [PDF]. The report cites a survey of about 11,000 executives and shows that concerns about the impact of AI on employment have attracted some attention from business leaders.
Almost half of organizations expect to reorient their business models on new AI-driven opportunities
“Almost half of organizations expect to reorient their business models at new AI-driven opportunities (49%), while 47% will move their employees from the disruptive role of AI to other positions,” Job Report said. “Most employers plan to hire new people with relevant skills, but a key share (41%) also hopes to reduce the workforce as AI capabilities expand to replicate roles.”
Changes are expected. “On average, workers can expect two-fifths (39%) of their existing skill sets to be transformed or outdated over the course of 2025-2030,” Jobs Report said.
As AI tools expand, workers need to learn new skills, but that has always been when technology shapes society.
Adopting AI doesn't necessarily mean that a wide range of jobs will disappear. As IBM CEO Arvind Krishna recently told The Wall Street Journal, “While we've done a huge amount of work within IBM by leveraging AI and automation in certain enterprise workflows, our total employment has actually increased.
It should be noted that to the extent that AI is creating new jobs at IBM, the company's current job listings show roles that appear outside the US in low-wage regions like India.
Scholars are concerned, but not within the scope expressed by Amodei.
In a recent paper entitled “Artificial Intelligence and the Labor Market,” economists Menaka Hampole (Yael), Dimitris Papanicolau (Northwest), Lawrence DW Schmidt (MIT) and Brian Siegmiller (Northwest) found that AI is limited to labor market adaptation because of labor market adaptation.
“Overall, we see a suppressed effect of AI on employment due to offsetting effects. Highly exposed occupations experience relatively low demand compared to less exposed occupations, but the resulting increase in productivity productivity increases overall employment across all occupations,” the authors conclude.
Co-author Papanicolau, a professor of finance at Kellogg School of Management at Northwestern University, Register It's difficult to predict the future, and easy to predict what's already happening.
“In the paper, what we're doing is watching past episodes of what people called AI in 2010. This is essentially machine learning or big data, or inherently predictive. “So I think that's very different from what humanity is doing now.”
Papanikolaou said the goal of the study was to draw out a distinction between the meaning of job exposure to AI and the meaning of labor demand for that particular job.
“People are thinking that when their work is exposed to AI, it means that AI is trying to do my job, then I'm going to finish the job,” he explained. “So the point we're trying to do is that it's not necessarily the case. There are some potentially offsetting factors.”
What's important is how AI can do all the tasks involved with a particular job. If AI is all okay, labor demand could decrease. But if AI is superior in some aspects of a person's job and is poor with other responsibilities, it's not too bad.
“It could act as a countervailer, because it essentially means that workers can free up time to concentrate on tasks that are not exposed,” he said. “In a nutshell, what's important is not just the average exposure of your job, but also how it's distributed across the various tasks you're doing.”
As reported last month, economists Anders Humram and Emily Vestergaard found that in a paper entitled “Large Language Models, Small Labor Market Impacts,” AI chatbots had little impact on 11 different occupations in Denmark in 2023 and 2024.
Even the most transformative technologies, such as steam power, electricity, and computers, took decades before they had a massive impact on the economy.
Asked about Amodei's concerns, Humlum stated, “Dario Amodei's predictions are thought-provoking and certainly worth considering. After all, he has one of the best views about the technology ahead of us lying.
“But we've now been experiencing two and a half years with AI chatbots that are spreading widely in the economy. And looking at the data, these tools aren't actually making a big difference in job employment or revenue for our professions.
Furthermore, looking at a broader history, even the most transformative technologies, such as steam power, electricity, and computers, took decades before they had a massive impact on the economy. Moreover, even when these technologies dramatically changed production, they did not cause massive unemployment in the long term. ”
There is so much uncertainty and disagreement about the long-term impact of AI on employment, why does Amody appear on such bold predictions? Well, from a marketing perspective, it's never painful to portray your product almost magically in its ability to change the world. It may also help to ensure he and the company are seated at the regulatory table. ®
