TLDR: Artificial intelligence’s humanistic shortcomings might not be intrinsic qualities of neural networks or silicon transistors but rather the result of the strategies we have used thus far to design LLMs. This goes way beyond grammar, nor is it about abandoning LLMs as a dead end. For neural networks to create on par with humans, we have to think creatively about how our brains are trained in things like taste and aesthetics, and then collect data, run experiments, and design model weights that approximate those routes and inputs, rather than starting with the last link in the chain— the written or artistic product. Instead, we could distill human intelligence itself by measuring things like inputs (books read) vs. outputs (books written). Such problems and questions can only be explored through deep collaboration between humanists and technologists.

Unfortunately, ChatGPT’s writing deficiencies have become some of its most derided qualities. But fixing style or getting rid of em dashes are cosmetic changes that might improve writing but do not address much deeper, more complex, and more interesting questions about how human brains work in humanistic dimensions shared by all people, not just humanities majors.

I am very interested in the intellectual challenge of making LLM writing and visual generation better than it currently is, whether through improving user interfaces to allow for different preferences and better results, or by deeper structural and systemic changes to how GPTs use their neural networks to manipulate and produce language. Doing so would pay dividends not just in better outputs, but in better training of models in the fundamental way their nodes and weights function in all domains, not just in writing. Those who say that LLMs can never achieve human-level thinking might not be thinking deeply or creatively enough about how humans think.