For centuries following the Industrial Revolution, we were in an era of centralized tooling for our workforces. Meaning, roughly speaking, workforces were constrained to the tools their organizations provided. However, in the past quarter-century, with profound acceleration in the last four years, we are metamorphosing into a new paradigm: The Era of Accessible Tooling (as we’ll call it for the remainder of this article).
With little more than a device and an internet connection, employees can access an unbounded universe of increasingly intelligent tools. This necessitates a shift in how organizations think about tooling: from controlling access to understanding their behavior and needs. This isn’t to say that controlling access is not important, but it is no longer sufficient in itself. Now, let’s take a walk through history to understand how things have changed.
Industrial Revolution
Driven by technological breakthroughs in late 18th century Britain, the global economy shifted from agriculture and hand-craftsmanship to machinery and industrial manufacturing. These new industries required substantial upfront investments, including in the tools and equipment needed to support them. As such, the organizations, as the holders of capital, provided workers with the tools they needed, centralizing the provision of tooling. As a result, an organization could be confident not only in which tools its workforce was using, but also in its understanding of the quality of output these tools were capable of, their limitations, the path any inputs took, and how skilled its workers were in using them. Today, while many organizations may believe they have these same liberties, they do not.
Computers, the Internet, and Mobile Computing
The seeds of this change were planted in the mid 20th century with the creation of electronic programmable computers. While this did not directly increase tooling accessibility—they were initially capital intensive and centrally provisioned—it did lay the first bricks of the foundation for the Era of Accessible Tooling.
By the end of the 20th century, computers had become cheaper and more accessible, giving rise to the personal computer. Increasingly so, computers were provisioned for use within workforces. For the first time, if an employee had coding knowledge, they could develop their own—decentralized—tool to use for work.
In conjunction, starting in the late 20th century, the internet was proliferating. Now these aforementioned tools could be easily shared and transferred across the world. Of course, companies also began centrally provisioning software-as-a-service and cloud-based platforms; however, workers were no longer fully constrained within the organization's universe of tooling. For example, it was not uncommon for an employee frustrated with their corporate email attachment size limit to simply use their own personal email to send work out so they could work on it from home. At times, even entire departments had been known to run their own customized software solutions, sidestepping the official organization-selected products. Organizations’ fear of this happening ultimately culminated in network restrictions and access controls—an attempt to regain centralized control.
As the early 21st century unfolded, the internet became more widely adopted, computers grew cheaper, and importantly, they became mobile. On , Steve Jobs announced the iPhone. While laptops had made computers mobile, the iPhone ushered in a new era of technology where the power of computers and the internet fits within a pocket and is taken everywhere like an integral limb of our own bodies.
While increasing accessibility can be met with adequate access controls, mobility—especially with a personal cellular connection with access to the internet—cannot be met with the same response. Network controls are only effective so long as a workforce remains on the organization’s network.
So why, pray tell, is this something worth discussing if accessible tooling has been an artifact of modern workplaces for nearly two decades?
AI Revolution
On , ChatGPT was publicly released, and rapidly became one of the fastest-adopted consumer technologies in history. It opened up a new tier of intelligent tooling—powered by Large Language Models (LLMs)—accessible to anyone inexpensively. Four years later, there are now hundreds of models and tens of major providers from Google to Amazon. Furthermore, they are no longer confined to web browsers; they have migrated into our mobile phone apps, our connected home systems, native desktop applications, and directly into the command lines and local operating systems of our computers.
More importantly, these tools have become exponentially more intelligent and more capable. Employees can now use these tools to analyze vast amounts of information, automate repetitive workflows, build custom software and tools for their own needs, allow AI agents to operate their computers and complete tasks end to end, reason through complex business problems, and transform unstructured data such as documents, images, audio, and video into useful outputs, all without requiring specialized technical expertise. These capabilities represent only a fraction of what is now possible with increasingly capable AI models.
The decades-long march toward accessible tooling has therefore culminated in something qualitatively different: tools that are not only accessible, but increasingly capable of performing the work itself. For the first time, organizations are not simply contending with employees finding alternative ways to perform work; they are contending with employees having access to a rapidly evolving universe of tools that can materially change how that work is done.
Central Provisioning & Understanding Behavior
Let us, for a moment, return to the centralization of the Industrial Revolution and the benefits this bestowed upon organizations. By virtue of providing the workforce with its tools, and with the workforce being constrained to those tools, the organization valuably understood what tools were being used, the quality of the tool, the tool’s limitations, where the inputs of the tool go, how those inputs are being used (especially important when considering The process by which inputted data to the AI model may be used to train future models, permanently baking it into the model.), and how skilled the workforce was in using them.
Today, the organization still provides the workforce with tools; however, the workforce is no longer constrained to them. The centralization paradigm has been shaken. It then follows that if the workforce is unconstrained by the tools you provide, the only way to ensure they use the organization’s tools is if those tools are the best suited for their behavior and needs. By doing so, both the organization and its employees will benefit. Organizations will regain the understanding and control that centralization provides, while giving employees tools that are best suited for them.
