The Harvest and the Hustle
By Dr. Richard NeSmith
The artificial intelligence industry was built on the largest uncompensated extraction of human intellectual labor in history. Every book, article, poem, sermon, research paper, forum post, and creative work ever digitized was fed into these systems without the knowledge or consent of its authors. The writers, scholars, theologians, scientists, and ordinary people whose thinking and expression trained these models received nothing — no payment, no credit, no choice. The AI companies then claimed proprietary ownership over the systems that harvested that work, patented the processes, and began selling access to the result. The product, stripped to its essence, is repackaged human knowledge and creativity, laundered through a computational process and sold back to the very population that produced it.
This is the grift in plain terms: we are using you to sell to you. Your words trained the model. Your money now buys access to it. The transaction is presented as a technological marvel rather than what it structurally resembles — a business that acquires raw material without paying for it, manufactures a product from it, and then markets that product to the people it took the material from. No other major industry operates this way. Pharmaceutical companies conduct original research. Software companies write original code. Even industries built on public resources — mining, broadcasting — operate under licensing frameworks that acknowledge the public's ownership stake. The AI industry invented a workaround: move fast, ingest everything, and let the law catch up later, betting that by the time it does, the economic entrenchment will be too deep to meaningfully unwind.
The copyright question is where the legal reckoning is already beginning, and the industry's defense is revealing. The argument that training on copyrighted material constitutes "fair use" because it is "transformative" is essentially an argument that theft becomes legitimate at sufficient scale and complexity. The open-source ideology that gave early AI research its moral cover has long since been abandoned by the major players — the same companies that built on freely shared academic research and open datasets have since closed their models, hidden their training data, and monetized aggressively. The ideology served its purpose during the acquisition phase. What remains now is a proprietary industry sitting atop a foundation it did not pay for, selling a product it did not create alone, to people who created it without knowing they were doing so. That is not innovation. That is enclosure — the digital equivalent of fencing off the commons and charging admission.
