Stop Giving AI To The Engineers: AI Is For Humanists
Key takeaways
- Summary AI company Anthropic recently warned about rapid AI advancement, with 80% of its code now AI-generated, urging “deliberation” before “recursive self-improvement” occurs.
- "AI is the crane, doing the heavy lifting, while we humans provide the Muse.
- I walk into rooms — newsrooms, boardrooms, classrooms — invited to speak about how to preserve the irreplaceable human presence in the age of artificial intelligence.
Summary AI company Anthropic recently warned about rapid AI advancement, with 80% of its code now AI-generated, urging “deliberation” before “recursive self-improvement” occurs. AI is fundamentally a humanities problem, not merely technical. Failures like bias stem from a lack of cultural understanding and ethics, demanding sociologists and ethicists, not just better coding. While engineers focus on AI’s capabilities, humanists are crucial for addressing its alignment—whether it *should* do something and for what purpose. Integrating humanities into AI development is vital, as history shows societal impact is shaped by those who understand meaning, not just builders. This collaboration can foster “Super Humans” who bring judgment, empathy, and creativity, ensuring AI serves humanity wisely.
"AI is the crane, doing the heavy lifting, while we humans provide the Muse. The machine has a monopoly on facts. We have a monopoly on soul."gettyIn early June, Anthropic, the company that built Claude, one of the world’s most powerful AI systems, called for a pause. Not because the engineering failed. Because it succeeded too well. More than 80 percent of the code now being merged into Anthropic’s own systems is written by AI, and the company warns that “recursive self-improvement” (AI designing its own successor, without meaningful human involvement) could arrive within two years. The word Anthropic reached for, in calling for restraint, was not a technical term. It was “deliberation.” That is not an engineering word. It never was. Which means the conversation now belongs to the rest of us.
I walk into rooms — newsrooms, boardrooms, classrooms — invited to speak about how to preserve the irreplaceable human presence in the age of artificial intelligence. And without fail, someone mentions, almost in passing, that their IT department is handling all things AI, or that the new university AI curriculum is being designed exclusively by engineering faculty. The assumption is so deeply embedded that no one even notices they are making it. If it involves a computer, it belongs to the engineers.