"Something Big Is Happening" ("Something Big Is Happening") - with that post on the X (X) network, Matt Schumer recently raised the alarm. Matt Schumer is an American technology entrepreneur and investor, co-founder and CEO of OthersideAI. He is considered an influential voice in the field artificial intelligence.
His post went viral and has been viewed more than 80 million times so far. The message is alarming: a big shock is coming very soon labor market caused by artificial intelligence. According to Schumer, whoever does not act now will hang, he writes DW.
What jobs does Schumer see as at risk?
The impact of artificial intelligence will be "significantly greater" than during the covid pandemic, Schumer is convinced. Within one to five years - "probably sooner" - no computer-based work today will be secure.
He cites the areas of software development, law, finance, medicine, accounting, consulting, content creation (PR and journalism), and customer support as being particularly affected.
"If your work is done on a screen — if it essentially boils down to reading, writing, analyzing, making decisions, and communicating via keyboard — much of it will be taken over by artificial intelligence. Not in some distant future, but now. That process has already begun," Schumer writes.
The question arises, however, whether Schumer's short essay is really a sober analysis of technological change – and therefore a serious warning to employees.
"It sounds dramatic, but I understand why. I am experiencing these changes myself. We are a two-woman company and with the help of artificial intelligence in many areas we achieve the same performance that previously required a team of 20 employees," says business advisor and artificial intelligence expert Claudia Hilker in an interview with ARD.
Critique of Schumer: Panic and problematic assumptions
Back in the spring of 2025, Dario Amodei, CEO of Anthropic, predicted that artificial intelligence would make about 50 percent of all entry-level office jobs redundant by 2030. Anthropic is the maker of Claude, one of ChatGPT's main competitors.
Schumer's warning about mass layoffs is not new, so it's not new. However, it has drawn sharp criticism. Fortune magazine's technology analyst Jeremy Kahn accuses Schumer of "spreading panic."
Artificial intelligence researcher Gary Marcus shares a similar view, saying: "I think we should start from the facts and not simply spread fear."
At the center of criticism are the assumptions on which Schumer bases his claims. Namely, he starts from the fact that the latest models of artificial intelligence - specifically GPT-5.3 Codex from the OpenAI company and Opus 4.6 from the Anthropic company - possess the ability of real "reasoning". In other words, to mimic human logical thinking, rather than just calculating the statistically most likely answer based on the training data. According to Schumer, those models "don't make mistakes anymore."
However, this assumption is questionable. Schumer, critics point out, is selective in the studies he cites. He omits recent research from the California Institute of Technology (Caltech) and Stanford University, which shows that so-called grand language models still make serious logical errors and show weaknesses in coherent reasoning.
Warnings also come from Apple
Previously, Apple showed in its study The Illusion of Thinking that models based on "reasoning" practically collapse and produce numerous errors when tasks are highly complex.
In addition, a study by the American software company Black Duck points to serious security risks: the problem with code generated by artificial intelligence is often not in obvious errors, but in the "illusion of correctness".
By ignoring these findings, Schumer, according to experts, underestimates the limitations of today's models: they still make mistakes, "hallucinate" and are only reliable to a limited extent.
Routine jobs are under attack - not entire professions
The conclusion that entire jobs will disappear in the short term due to artificial intelligence is therefore considered exaggerated. Most experts expect that AI will primarily strongly automate individual parts of jobs.
"Knowledge-intensive, digital and standardized activities are particularly affected. But not entire activities will disappear, but primarily routine tasks," emphasizes Hilker.
Strategy, leadership, responsibility for decisions and creative conception will gain importance.
"Artificial intelligence can analyze, but it doesn't bear responsibility. It doesn't build trust and it doesn't take legal responsibility."
A revolution of efficiency, not mass unemployment
Schumer's post is not only a warning - it can also be read as a promotional text, because the author directly invites readers to subscribe (for money) to the latest artificial intelligence models in order not to fall behind in the labor market.
Hilker sees it differently: "We're not facing mass unemployment. We're facing a radical efficiency revolution. The key question is not whether artificial intelligence will take over my job, but whether I'm already using it professionally enough to be one step ahead in my industry."
Her conclusion is: artificial intelligence does not replace humans. But people with AI knowledge replace those without.
Source: DW
