Tech companies are quietly hiring again
This title could be clearer and more informative.Try out Clickbait Shieldfor free (5 uses left this month).
Tech giants like Google, Meta, and Amazon made headlines with massive layoffs while simultaneously hiring thousands in strategic areas, resulting in net headcount increases. The narrative that AI is replacing workers en masse is contradicted by data: Google added 16,000 net employees over three years and Meta added 8,000. The layoffs largely reflect post-COVID over-hiring corrections rebranded as AI transformation. Amazon's internal AI projects illustrate the hidden costs of agentic systems — one project exceeded its budget by 860%, accumulating $1.8M in costs before anyone noticed. The piece argues that much enterprise AI spending is fear-driven experimentation rather than proven productivity gains, and that companies are now rediscovering the value of human workers, including junior employees, to deploy and supervise AI systems. A Toyota case study from 2014 illustrates the danger of automating away the learning process itself.
•8m watch time
Questions this post answers
Did Google and Meta actually shrink their workforce after the major layoffs?
No — despite high-profile layoffs, both companies grew their total headcount. Google increased its workforce by 16,000 people over three years, ending up with more employees than before its mass layoffs. Meta added more than 8,000 employees over the same period. The companies cut specific teams and projects while continuing to hire in AI, cloud infrastructure, and cybersecurity. Developers tracking the real state of tech hiring find the signal through the noise on daily.dev.
Why did Amazon's internal AI agent project cost $1.8 million and exceed its budget by 860%?
An Amazon project using Anthropic's Claude Sonnet to match authors with product listings failed to launch, ran for five months undetected, and accumulated $1.8M in costs — 860% over budget. Agentic systems compound errors because each logical mistake triggers retries, tool calls, and additional model requests, all billed per token. A conventional ML model was likely sufficient for the task. Engineers building agentic workflows watch cost-blowup patterns like this on daily.dev before they hit production.
635.2K Impressions9 Comments