What Is AI For? A Question as Old as Writing
Artificial intelligence may be less like the internet and more like the invention of writing, externalizing portions of human reasoning just as writing externalized memory, forcing society to reconsider not only what AI is for, but what human beings are for.
What is AI for?
The question appears simple. Most people answer it without hesitation. AI helps us work faster, automate routine tasks, analyze information, and generate content. Yet the more closely one examines the question, the less satisfactory those answers become.
After all, what is the internet for?
Thirty years ago the answer might have been email, research, communication, or commerce. Today the question feels almost impossible to answer because the internet became something larger than a tool. It became part of the infrastructure of modern life. We communicate through it, learn through it, shop through it, work through it, and increasingly govern through it. Its original uses still matter, but they no longer define its significance.
Artificial intelligence may be following a similar path. Yet the deeper historical comparison may not be the internet at all. It may be writing.
When Writing Learned to Remember
That claim sounds exaggerated until one considers what writing actually accomplished. Before writing, knowledge lived primarily within human minds. Memory, oral tradition, storytelling, and direct instruction carried information across generations. To know something required that someone remember it. The scale of civilization was therefore constrained by the scale of human memory.
Writing fundamentally altered that relationship. For the first time, information could exist independently of the human brain. A clay tablet could preserve a harvest record. A scroll could preserve a law. A manuscript could carry ideas across centuries. Humanity had effectively invented a way to externalize memory.
Not everyone viewed the development positively. In Plato's Phaedrus, Socrates recounts a story in which writing is criticized because people may come to rely on external symbols rather than internal recollection. Knowledge would appear to increase, critics feared, while genuine understanding might decline. The concern sounds remarkably familiar. Many contemporary arguments about AI reflect similar anxieties regarding cognitive dependence and intellectual atrophy.
History ultimately demonstrated that writing was neither purely beneficial nor purely harmful. It reduced the need for memorization in many contexts, but it also enabled philosophy, science, literature, law, accounting, and large-scale civilization itself. Writing changed how human beings thought, learned, and remembered. The technology did not diminish humanity. It transformed humanity.
Artificial intelligence may represent a similar shift.
If writing externalized memory, AI appears to be externalizing portions of reasoning and synthesis. Researchers can now explore vast bodies of literature in minutes rather than months. Students can receive explanations and feedback on demand. Professionals can delegate routine drafting, summarization, and analysis to systems operating at extraordinary speed. The technology does not replace human thought, but it changes the relationship between human beings and certain cognitive tasks.
That possibility helps explain why the debate surrounding AI feels unusually consequential. Society is not merely adopting another tool. It may be renegotiating the boundary between internal and external cognition. The invention of writing shifted part of memory outside the mind and into the world. Artificial intelligence may be shifting portions of reasoning, pattern recognition, and synthesis into systems that are available whenever needed.
Image generated using OpenAI's DALL·E image generation model in response to the prompt: "What is AI?" (2026).
What Happens When Thinking Changes?
History suggests that major cognitive technologies rarely produce simple outcomes. Writing reduced the importance of memorization while expanding the possibilities for complex thought. The printing press reduced the scarcity of books while transforming religion, science, and education. The internet dramatically increased access to information while reshaping communication and attention. Each technology expanded human capabilities while simultaneously changing what society valued.
Artificial intelligence may force a similar reassessment. If information becomes abundant and inexpensive, the relative importance of judgment may increase. If routine analysis becomes automated, interpretation may become more valuable. If AI can answer questions instantly, the ability to ask meaningful questions may become the scarcer skill.
That possibility carries an unexpected implication for education. For generations, schools and universities have devoted considerable effort to helping students acquire information. As AI systems become increasingly capable of retrieving, summarizing, and explaining information, educational institutions may find themselves focusing more heavily on judgment, wisdom, interpretation, and the ability to navigate competing ideas. The challenge may shift from helping students find answers to helping them understand which questions are worth asking.
Seen in this light, many contemporary debates about AI are not really debates about technology. Discussions about alignment become discussions about values. Conversations about governance become discussions about responsibility. Arguments regarding educational uses of AI become arguments about the purpose of education itself. Behind nearly every debate lies an implicit theory of what human beings are for.
That realization may explain why the question "What is AI for?" remains so difficult to answer. The problem is not that we lack sufficient knowledge about the technology. The problem is that the question ultimately points beyond the technology. Every civilization develops assumptions regarding the purpose of knowledge, the purpose of education, and the purpose of human life. The tools it creates inevitably reflect those assumptions.
The invention of writing changed what it meant to remember. Artificial intelligence may change what it means to think.
Further Reading
- Plato, Phaedrus
- Norbert Wiener, The Human Use of Human Beings (1950)
- Neil Postman, Technopoly: The Surrender of Culture to Technology (1992)