Companies are not ready for the amount of demand they’re going to see for computers. Today, IT teams buy employees computers and those employees tell those computers what to do. People log in, work, and sign out, all on one device. That’s all about to change with the influx of AI agents entering the workforce.
AI agents that control computers to get tasks done will soon outnumber the people who tell them what to do. Companies need to stop treating high-end computers as personal equipment and start treating them as shared infrastructure that employees and their agents can use whenever they need it. That’s why we’re building Phaze. It lets people and their AI agents work together and access powerful computers from anywhere, on any device, without sacrificing performance. When compute can live anywhere, the computer on your desk becomes optional.
We’ve spent more than a decade building remote desktop technologies. We were the founding team behind Parsec, which we sold to Unity in 2021. Now, we’re back with Phaze to help companies manage this growing demand for computing power. Phaze combines high-performance remote access technology with a machine management system, allowing companies to share powerful computers across employees and AI agents without requiring dedicated hardware for every user.
Computers will be optional
By computers, we mean the device on your desk. People who love computers will still build, mod, and tinker with them the same way car enthusiasts restore and customize their cars. For most people, however, a computer is just a tool. The CPU, GPU, RAM, and storage that get work done don’t need to sit in front of you. They can live in an office, a rack, a data center, or the cloud, and they can be available when a task needs them. And increasingly, those tasks are being done by people and their agents working together. Compute demand will come from both.
Companies need to think differently about hardware
Companies still operate as if a computer has to belong to an individual employee. Each engineer and designer gets a workstation, and each machine gets sized for the most demanding work that person might need to do. Companies buy powerful, increasingly expensive workstations for individual employees, even though those machines may sit idle for much of the day. That is especially true for work like CAD, video production, engineering, and game development, where specialized software requires high-end hardware. Companies end up buying for peak demand while using only a fraction of that capacity throughout the day.
At enterprise scale, the economics become bank-breaking. A large manufacturer can have tens of thousands of engineering workstations, often alongside laptops for the same employees. One we’ve spoken with has 30,000 workstations that can cost more than $10,000 each, with much of that computing capacity sitting idle throughout the day. That is more than $300 million of hardware purchased for peak demand rather than shared across the organization. And that doesn’t even account for power, cooling, maintenance, and replacement costs.
That model cannot survive what comes next. Compute requirements are climbing, hardware is getting more expensive, and AI agents are dramatically increasing the number of users that need access to computers. An employee may have several agents working at the same time, each capable of using software and requiring a computer of its own. Buying another dedicated workstation every time a person or agent needs more compute is economically unsustainable.
Turning workstations into infrastructure
This is where Phaze comes in. Instead of buying powerful computers and tying each one to a person, desk, or location, companies can make that compute available across the organization. IT can manage and assign machines as a shared resource, employees can access the hardware they need from wherever they are, and AI agents can use those same machines to get work done. When a person or agent finishes a job, that compute can return to the pool for someone else to use. IT teams can also deploy their own agents to troubleshoot and fix users’ machines.
The physical computer should not determine who can use the compute inside it. Phaze separates the two, so a workstation can serve whoever needs it, wherever they are, and return it to the pool when the work is done. Instead of provisioning hardware around individual employees, companies can provision it around actual demand. Over time, most people won’t need a dedicated high-end computer, and AI agents will get the compute they need the moment a task is assigned.
In Piers Anthony’s Apprentice Adept series, Stile lives in two worlds at once: Proton and Phaze. That’s where our name comes from. With Phaze, you and your agents can do the same, working across many machines at once with whatever compute the task needs.
If your team and its agents need better access to powerful computers, sign up here.