Jan 18, 2024 - Technology

"AI made us do it" is Big Tech's new layoff rationale

Illustration of a person at a computer with their face covered with a box with the text "Replace" highlighted by a cursor.

Illustration: Shoshana Gordon/Axios

Big tech companies are pointing to a renewed focus on AI as a justification for layoffs.

Why it matters: The new trend in tech company layoffs could be less about replacing workers with AI, and more about replacing workers with a smaller number of workers who are more skilled in AI, for now.

Driving the news: Last week and again this week, Google has laid off hundreds of workers to focus more on AI. Some of those workers were on advertising or hardware teams, but others were working on one of Google's original AI tools, Google Assistant.

  • And Google has more cuts to come, per a memo to employees reported by The Verge, with news of around 100 layoffs among YouTube business staff breaking yesterday. Google spokesperson Courtenay Mencini told Axios the changes are about efficiency and focusing on the company's biggest priorities.
  • Salesforce, Duolingo, and even AI hardware startup Humane have also announced layoffs or hiring freezes as a part of an effort to focus on AI.
  • According to Layoffs.fyi, 58 tech companies have laid off 7,785 employees already in Q1 2024.
  • Muddu Sudhakar, co-founder and CEO of generative AI solutions company Aisera, told Axios that he's seeing a "a huge displacement of white-collar workers" in basic software developer jobs as well as database administrator roles.

When company layoff announcements come in waves, spokespeople often give the same reasoning for layoffs. Last year the shared refrain was "the macroeconomic climate," and now many tech companies are saying layoffs are a result of a "renewed focus on AI."

  • Tigran Sloyan, co-founder and CEO of technical skills assessment platform CodeSignal, told Axios that these announcements could mean that tech companies are "reorienting resources" to spend more on bigger salaries for highly skilled engineers.
  • But he said, it could also just be "a nice way to manage the PR element" of layoffs. Or it could mean that the companies are fully automating jobs and replacing workers with AI.
  • Contractor cutbacks by language-learning software maker Duolingo this month came as the firm was able to handle more of their work with AI, per Bloomberg.

Between the lines: There are still postings for extremely high-paying AI engineer positions — like the $900,000 one at Netflix last year that went viral — but few workers have the skills required for those jobs.

  • Those massive salaries are shaped by limited supply and high demand for PhD-level researchers. "There are very few people who have gone through that much training, that much education, and that much skill building," says Sloyan.
  • According to a recent CompTIA report, posting for AI jobs with specialized skills represents more than 10% of all tech job postings.

The intrigue: Sudhakar told Axios that while AI salaries are on the rise, generative AI salaries only capture a small section of the AI market.

  • "There is going to be downward pressure on general AI salaries as automation of code generation becomes more prevalent in the enterprise, especially with tools like Github Copilot," says Sudhakar.
  • Mark Muro, a senior fellow at Brookings Metro, said there's an immediate need for talent with AI skills. But the number of hires doesn't come near to the size of the past year's layoffs, and the new positions are concentrated in a small number of tech hubs.

The big picture: Large language models pose real risks for software engineering jobs, Muro says.

  • He points to forthcoming work from Brookings Metro showing that "'computer' jobs —including software development — face generative AI exposure scores of 90% and more, meaning that 90% of their work tasks could be done in half the time using ChatGPT."

Editor's note: This story has been updated with additional details about Google's cutbacks and updated layoff numbers from Layoffs.fyi.

Go deeper