TL;DR
The author tested an AI trained on his past posts and found the results fell into an uncanny-valley inconsistency. He argues that the daily discipline of writing and the thinking it forces can’t be replaced by AI, which may produce useful lists of ideas but not the deep connections that come from doing the work yourself.
What happened
The author describes a conversation with a friend about an AI model trained on his previous blog posts. The experiment supplied the model with headlines and opening paragraphs from his 2025 pieces to see if it could produce the remainder in his voice. After comparing a handful of AI-generated drafts with the originals, he found the results superficially plausible but ultimately off: the model sometimes took arguments in different directions, asserted confidence where he would express doubt, or reached conclusions he wouldn’t. He rejected the idea of replacing his daily writing with AI, saying the practice itself—forcing him to think, keeping his craft sharp, and honoring a daily pledge to readers—is more valuable than mere output. He and a friend referenced Ezra Klein’s view that AI, so far, is useful mainly for light research or structuring data and that outsourcing core thinking risks losing what makes writing distinctive.
Why it matters
- Daily writing can be a practice that shapes thinking and craft in ways AI output does not replicate.
- Relying on AI to fill creative gaps may produce superficially plausible but fragile content.
- As AI lowers the barrier to producing text, original work may become a clearer differentiator.
- Summaries or automated research can miss the specific connections an individual writer seeks.
Key facts
- Someone trained an AI model on the author’s past posts as a research exercise.
- The experiment fed the model headlines and opening paragraphs from the author’s 2025 posts.
- The author compared a handful of AI-generated drafts with the original pieces.
- He observed an uncanny-valley effect: output looked fine at first but felt off on closer inspection.
- The AI sometimes veered in different directions, made opposite arguments, or sounded overly confident.
- The author says he writes daily to practice thinking, strengthen his craft, and keep a promise to readers.
- Writer Ezra Klein said he finds AI mainly useful for light research or structuring data.
- Klein warned that outsourcing core research or summarization can prevent the writer from making unique connections.
- The author argues AI can supply idea lists but that those can become weak filler rather than genuine creative progress.
- The author believes AI will make it easier to stand out if more people take shortcuts and lower overall quality.
What to watch next
- Whether AI models improve at preserving a specific writer’s decision-making and uncertainty — not confirmed in the source.
- Whether the author or similar writers will begin integrating AI into daily creative rituals — not confirmed in the source.
- How reader preferences change as AI-produced writing becomes more common — not confirmed in the source.
Quick glossary
- Uncanny valley: A phenomenon where something artificial appears almost human-like but contains small imperfections that make it feel unsettling or wrong.
- Fine-tuning: Training a pre-existing AI model further on a specific dataset so it better matches a particular style or task.
- Prompt: The input text or instructions given to an AI model to generate a response.
- Summarization: The process of producing a condensed version of a longer text that captures its main points.
Reader FAQ
Did the AI perfectly replicate the author’s writing?
No. The author found the AI’s output superficially plausible but often off in direction, tone, or argument.
Will the author replace his daily blog with AI?
No. The author said he wouldn’t use AI for that purpose because the daily practice is essential to his thinking and commitment to readers.
How does Ezra Klein use AI?
According to the source, Klein uses AI for light research and to structure data, but not for core creative work.
Does the author think reading full books is still valuable?
Yes. He suggests reading entire books may now be more valuable because summaries produced by AI can miss unique connections.
The Suck Is Why We’re Here January 2, 2026 Daily Blog On a catchup call, I told my friend Nick Wignall how someone had trained an AI model to write…
Sources
- The suck is why we're here
- “AI-Writing Sucks” Is a Stupid Lie — Here's What Most …
- How I roll as a writer in the AI tsunami
- Why AI Art Sucks – ISSUE 002 – by SAIDY
Related posts
- The Riven Diffs: Re-examining Riven (1997) Through Image Differences
- Assessing Every-Third-Day Versus Daily Low‑Dose Aspirin: Comparative Benefits
- The Suck Is Why We’re Here: On Writing, AI and Creative Work