In the AI field right now, there’s a certain kind of cognitive dissonance at play. The same executives who are announcing the release of billion-dollar models are also warning in a quiet way that these models could eliminate white-collar jobs within ten years. Dario Amodei of Anthropic once said that AI could take away half of all white-collar jobs. It was warned by Sam Altman that whole types of entry-level work were in danger. Both of them have since taken back what they said. If you want to call it that, the damage was already done.
A computer science professor at Georgetown named this behavior “doom trolling” not long ago. The idea is pretty clear. In it, they talk about how big AI developers solemnly warn about the bad things their technology could do while still making and selling it. Newport said that their stance was one of helplessness, a type of acting concern that doesn’t really slow anything down.
That frame is worth thinking about for a moment because it explains something that’s been hard to say about the present. The fear doesn’t come from outside the business. It’s being made by people in the industry and comes in a nice package with the pitch decks and business subscriptions.
The effects are going to a real place. It looks like younger workers are stuck in an exhausting loop. Computer science students are told to learn AI tools so they can stay competitive, but at the same time, they are reading news stories that say those tools could make their jobs obsolete before they even start. That kind of tension is hard to get rid of. It often turns into low-level anxiety that people bring with them to job interviews, entry-level jobs, and their general ideas about what work will be like in ten years.
That being said, this is no longer just anxiety for HR leaders. Ravin Jesuthasan, who is in charge of transformation work at Mercer, made it clear that the organizations using these tools are the ones who need to take action. He said that the warning from the biggest names in AI wasn’t always wrong; it depended on what happened. Mass displacement doesn’t have to happen. But it is possible, especially if businesses see these tools as nothing more than replacements for people who work there instead of chances to rethink how work is organized.
That difference is more important than it might seem at first. There’s already proof that some businesses are using AI to hide layoffs that were going to happen anyway. Jesuthasan called this “AI washing” of earnings reports. It’s not just a communication issue when people blame automation for layoffs when other business pressures are really to blame. It adds to the general feeling of fear that makes planning the work force harder for everyone, even the companies that are letting people go.
It’s interesting that investors seem to have moved past the panic and are now acting with more reason. Mercer’s 2026 Global Talent Trends report, which is based on the opinions of almost 12,000 investors, executives, and HR leaders around the world, says that 97% of investors would look negatively on companies that don’t adopt skills-based, agile talent models. That’s not a worry about AI taking away jobs. This is a worry about businesses that can’t change fast enough. From that point of view, the fear has turned around.
It seems like the conversation about AI in the workplace has become stuck between two extremes: disaster and dismissal, with not much useful ground in between. It’s not more interesting to ask if AI will take away jobs. It depends on whether the companies that use it are ready to revisit those jobs and think of new ways they could be done. That kind of planning doesn’t get news coverage. It does not spread. But that’s probably what matters most, especially for the workers who care less about the risk to civilization and more about whether their job will still exist in three years.
Horrifying people with doom makes for great news stories. Still, it’s not clear if it makes for good leadership.
