'AI' is Not a Tool

In the realm of artificial intelligence, a prevailing sentiment asserts that AI is merely a tool, a means to an end. However, this viewpoint may be overly simplistic, as there are significant differences between AI and traditional tools. By examining these distinctions, it becomes clear that AI should not be considered a tool in the conventional sense.

First and foremost, tools are implements that individuals wield, utilizing their skills, expertise, and control to accomplish tasks. They provide sensory feedback, allow for continuous direction, and build expertise over time. Tools are predictable, reliable, and deterministic, with a bounded function that enables users to learn about the world, the task, and the underlying principles at play.

On the other hand, AI systems like large language models (LLMs) operate on a fundamentally different basis. They are characterized by automation, lacking the control, sensory feedback, and intention that define traditional tool use. AI systems do not require human expertise or continuous direction; instead, they autonomously act upon user requests. This request/automation cycle lacks the feedback and control present in tool use, and it does not facilitate learning or the development of expertise.

Moreover, AI systems are not wielded by humans in the same way that traditional tools are. Rather, they act on their own, often rendering human input obsolete. While some may argue that AI is simply a tool for reading and summarizing research papers, the truth is that it does not perform these tasks in the same way a human would. AI systems do not provide learning opportunities, help users formulate arguments, or contribute to the development of critical thinking skills.

In essence, AI systems, such as LLMs, are automation tools that differ significantly from traditional tools. They possess intrinsic properties that set them apart from implements used by humans, making it a misconception to equate AI with a mere tool. Recognizing these differences is crucial to understanding the implications and limitations of AI in our lives.