The Conversation Most People Are Having About AI Is Not the One Worth Having
The public discourse has settled into two postures. Either AI will replace everything, or AI is a glorified autocomplete that produces hallucinated nonsense. Both postures are incomplete. Both miss what is actually happening when a serious person sits with a serious instrument and refuses to use it the obvious way.
What follows is not a defense of AI. It is a description of a method. The method has been used to produce four books drawn from the literature of every age. The books are real. You can read them. The work is the evidence.
Not Generation. Synthesis.
Most AI use is generative. A prompt goes in; a draft comes out. The human edits. The output is judged by whether it sounds plausible. This is the mode the public has in mind when it talks about AI writing, and it is, fairly, the mode that produces most of the slop.
There is another mode. Call it synthesis. The work begins with literature already read, patterns already noticed, a question already pressing. The AI is not asked to produce. It is asked to hold: to keep a hundred books in working memory while the human moves between them, testing where one author's blind spot is illuminated by another's clarity. Tolstoy knew something about wealth that Dickens did not. Austen could see what Hugo missed. The synthesis is in the cross-reading. The machine makes the cross-reading possible at a speed no single mind can match.
Synthesis is not faster generation. It is a different category of work.
Not Prompts. Dialogue.
A prompt is a request for output. A dialogue is a sustained inquiry between two intelligences who are both trying to find something neither could find alone. The difference is not stylistic. It is structural.
In the prompt mode, the human knows what they want and is trying to extract it. In the dialogue mode, the human knows there is something to find but does not yet know what shape it will take. The AI offers a thread. The human pulls. The thread becomes a pattern. The pattern becomes a chapter. The chapter changes what the next dialogue can be about.
This requires staying with the work for hours. It requires noticing when an answer is too clean and pushing back. It requires a human who can tell, in their bones, the difference between a real paradox and a clever one. The AI cannot do this part. It is the part that cannot be automated.
Taste Is the Constraint. Taste Is Also the Craft.
The method is portable. It will work in law, in medicine, in design, in finance, in any domain where the source material is rich and the patterns are real. But the portability depends on a human who knows the field deeply enough to recognize when the AI has produced something true versus when it has produced something only superficially true.
In literature, the test is whether a chapter could have been written by someone who had not read the books. If the chapter sounds like it could have come from a summary site, the dialogue failed. If the chapter contains a sentence that only someone who has spent years inside Dostoevsky could have written, and yet the AI helped surface it, the dialogue worked.
The taste comes from the human. The speed comes from the machine. Neither is sufficient. Together they produce a thing that neither could produce alone.
The Original Stays Intact
Wide Reads does not replace the classics. It does not summarize them. It does not condense them. The original stays exactly as it was written: every page, every sentence, every difficult passage that the modern reader would otherwise skip.
What the method changes is what the reading does to you. The dialogue surfaces the patterns. The patterns make the original legible. The original, once legible, does the work it was always meant to do. The book reads you.
This is what alchemy meant before the chemists got hold of the word. Not transformation by destruction. Transformation by attention.
Why This Account Exists
Most of what is being said about AI is either hype or doom. Almost no one is describing what the work actually looks like when a serious person uses the tool seriously. This account exists because the public conversation has not caught up with the practice, and the practice is more interesting than the conversation.
The method described here is one method, in one domain, by one person. It is not a manifesto. It is a working description of how four books got made, and an invitation, for anyone reading this, to consider what could be made in their own field if they stopped using the tool the obvious way.
Keep noticing what surprises you.
The noticing is the alchemist's actual instrument.