Tech jobs have hit an AI air pocket. Is it temporary or permanent?

• Tech job cut announcements have picked up again recently.
• Some factors pressuring tech jobs could be temporary, while others may be permanent.
• How long will this "AI air pocket" for tech jobs last?

Multiple forces have converged to trigger a wave of job cuts across the tech industry. Some of these changes could be temporary, while others could stick and change the character of tech work forever.

The latest news feels bad. Meta is cutting another 8,000 jobs. Microsoft offered buyouts to long-term employees. Oracle shed thousands of roles. Block chopped 40% of staff.

These cuts are painful and stressful for the people involved. However, it pays to put recent numbers in context, gauge how much AI is to blame, and consider how permanent this situation might be.

A recent report from Challenger, Gray & Christmas found that US tech layoffs were worse in 2023, only three years ago. Compared to the Great Recession of 2008 and 2009, this feels like a blip. I remember. I was there. It was terrifying.

Here are some potentially temporary sources of pressure on tech jobs, along with others that could be permanent.

Temporary pressure points

• Tech companies dramatically over-hired in 2020 and 2021 during the pandemic boom. They are still adjusting. Once that process is complete, job cuts could slow or even reverse. There's still demand for tech workers. Job postings for software development roles have risen sharply this year and have been climbing since late 2022, when ChatGPT launched. The supply of software engineers has risen dramatically, though, making the current situation feel worse, especially for some new graduates.
• Some tech giants are investing heavily in data centers and other AI infrastructure to develop new models and related services. That AI capex boom could peak around 2028, according to analyst projections. After that, some of these companies may have more financial flexibility, which could support a pickup in hiring.
• There is also what I'd call a potential "AI air pocket." Companies are experimenting with automation tools and pausing hiring as they evaluate their effectiveness. While AI has been blamed for job losses, it accounted for only about a quarter of cuts this year, according to Challenger estimates. As experiments conclude, some roles could return, especially where automation proves imperfect.
• Tech job losses are "not a result necessarily of AI successes," Gartner senior director analyst Kathy Ross said. "Instead, the layoffs seem to be part of a broader strategy to reinvest funds in AI, hoping for success down the line."

Permanent job market changes

• Some types of jobs could disappear for good, particularly in areas like customer service, where AI systems can handle routine tasks more cheaply and efficiently. If you handled horses when the automobile arrived, then you probably lost your job. Should we have all kept riding horses instead of using cars and continued moving around slowly? Probably not.
• AI also increasingly rewards employees who build, create, and sell, rather than those focused on process and oversight. In Silicon Valley, that shift is showing up in flatter organizations, fewer middle managers, and a growing preference for employees who identify as "builders." AI tools can handle some routine work — analyzing internal data, coordinating workflows, or managing repetitive tasks — reducing the need for roles centered on supervision or process maintenance. By contrast, workers who produce tangible output can potentially use AI to get more things done more quickly. Comfortable roles that relied on maintaining and repeating established processes are being exposed as less essential.
• New kinds of tech jobs will emerge, and some of these may become long-term professions. AI tools like Claude Code are enabling a new "Design Producer" role in Silicon Valley, where senior individual contributors focus on creating, guiding, and elevating work, using judgment and taste to amplify team output, rather than managing people through traditional hierarchies. Then there are "robot wranglers," who move, fix, and train AI-powered robots.

Sign up for BI's Tech Memo newsletter here. Reach out to me via email at abarr@businessinsider.com.