The layoffs announced by some of the world’s largest tech companies in recent months have sparked an intense debate about the impact of artificial intelligence on employment. However, behind the headlines, there is a much deeper shift: companies don’t seem to be cutting their investments, but rather shifting them from personnel to the infrastructure that powers AI. GPUs, data centers, high-speed networks, and storage are taking the place that, for decades, was occupied by workforce expansion.
The key to the new growth model in 30 seconds
- Large tech firms like Oracle, Meta, and Microsoft are maintaining record investments despite announcing staffing adjustments.
- Spending is shifting toward data centers, GPUs, networks, and memory for artificial intelligence.
- Since 2020, revenues continue to grow strongly, while employment evolves at a much slower pace.
- AI not only automates tasks: it changes how a company can increase its productive capacity.
Recent announcements illustrate this trend well. Oracle, Dell, Cisco, Meta, Microsoft, Atlassian, and GitLab have announced reorganizations involving thousands of layoffs or reassignments of staff. At the same time, these same companies, or their main competitors, are allocating historic amounts to building infrastructure for AI.
Meta projects investments exceeding $100 billion, Oracle is around $90 billion, and Amazon continues expanding its global network of data centers to support the growing demand for AI services.
At first glance, it might seem like cost-cutting strategies. However, the numbers point in a different direction.
Growth no longer depends solely on hiring
For much of Silicon Valley’s history, there was a relatively direct relationship between revenue and employment. When a company doubled its business, it typically needed to hire more engineers, support staff, salespeople, and operations teams.
That pattern is beginning to break down.
Data aggregated from companies like Microsoft, Amazon, Alphabet, Meta, and Apple show that until around 2020, revenue and workforce growth moved quite similarly. Since then, both curves have started to diverge.
Revenue continues to increase at an accelerated pace.
Workforces keep growing, but much more slowly.
This doesn’t mean companies need fewer people to operate. It means each new employee can generate more value thanks to intensive use of AI tools and automation.
GPUs become the new factor of production
The main difference compared to other technological revolutions is the destination of capital.
Just a few years ago, when a company wanted to scale capacity, a significant portion of the budget ended up funding new hires.
Today, an increasing share is invested in:
- AI accelerators like NVIDIA Blackwell or AMD Instinct;
- HBM memory and specialized DRAM;
- High-performance SSD storage;
- InfiniBand and gigabit Ethernet networks;
- New data centers with electric loads in the hundreds of megawatts.
In other words, computing capacity is beginning to become a strategic asset comparable to the human capital it historically replaced.
It doesn’t fully replace talent, but it alters the balance between the two.
Productivity shifts the equation
Artificial intelligence allows automation of some repetitive tasks in software development, technical support, document analysis, customer service, or content generation.
This doesn’t necessarily mean immediate job cuts.
What it does change is the expected productivity of each professional.
If an engineer develops more features, an analyst processes more information, or a salesperson spends less time on administrative tasks, the company can handle a larger volume of business without increasing staff at the same rate as before.
This partly explains why many companies report double-digit growth without corresponding increases in employee numbers.
Not all layoffs mean less talent
Another important aspect is that many adjustments do not solely involve reducing personnel.
Meta acknowledged that while it was cutting thousands of jobs, it was also reallocating a significant portion of employees to AI-related projects.
GitLab explained that part of its reorganization aims to fund infrastructure specific to workloads oriented around AI agents.
Nutanix recently announced a near 5% workforce reduction while increasing resources allocated to AI, Kubernetes, and data center modernization.
Rather than abandoning talent, what’s emerging is a shift in priorities.
The next battle will be over infrastructure
Over the past two years, the conversation has centered on AI models.
However, it is increasingly evident that the real competition is shifting to hardware.
The availability of GPUs, memory, electrical power, and data center capacity has become one of the main limiting factors for deploying AI at scale.
For this reason, multi-billion-dollar investments are no longer solely aimed at developing models but also at building the infrastructure necessary to run them.
This phenomenon affects chipmakers, cloud operators, storage providers, network companies, and data centers.
The labor market enters a new phase
The conclusion may not be that artificial intelligence directly destroys jobs.
What seems to be changing is the mechanism by which tech companies grow.
For decades, growth meant hiring.
Today, it also means deploying more GPUs, expanding data centers, incorporating high-performance storage, and increasing inference capacity.
Companies continue to invest billions of dollars.
They are just investing in different assets.
And this shift could ultimately redefine not only the tech sector but also how business growth is measured over the next decade.
Frequently Asked Questions
Is artificial intelligence replacing workers?
Not necessarily. Many companies are reorganizing teams and automating certain tasks, but at the same time increasing investment in new AI-related areas.
Why are tech companies laying off staff if their revenues are still growing?
Because part of the capital previously used to expand the workforce now funds infrastructure such as GPUs, data centers, networks, and storage needed to run AI workloads.
Which sectors benefit from this change?
Semiconductor manufacturers, cloud providers, data center operators, storage companies, networks, and energy firms are among the main beneficiaries of this new phase of tech investment.
Is this just a temporary trend?
All signs point to no. The investments announced for the coming years suggest that AI infrastructure will remain a strategic priority for major tech firms.
