Use any of this freely. The only thing asked is that you keep it honest, which is easy here, because the honest version is the interesting one. Do not call it the world's first or the definitive anything. It has not earned those words yet, and the fact that it refuses to claim them is part of what makes it worth sharing.
The problem it speaks to, in one breath
There is too much to read, you cannot tell who to trust, and almost every reading list you have ever seen was either someone's opinion or someone's affiliate income. Meanwhile the field itself has split in two, English and Chinese, and most maps only show you one half.
The turn
Someone built a reference library for the whole field that you can actually check. It ranks the texts, not the people. Every ranking shows its evidence. You are invited to prove it wrong, in public. It is free, it is built to include both the American and Chinese literature, and it sells you nothing.
A drop-in post you can adapt
Most “best AI books” lists are either someone's opinion or someone's affiliate link.
I just came across something different: The AI Canon. A free, public reference library of the texts that define AI, built on an open method you can actually inspect.
What makes it stand out to me:
- It ranks texts, not people. The thinkers and labs are described, never ranked against each other.
- You can check every judgment. The method, the data, and the audit files are public, and each ranking links to the evidence behind it.
- It invites you to prove it wrong. Disagree with a ranking? File a challenge with evidence. Every challenge and its resolution is published.
- It takes China seriously. It is built to include the Chinese-language literature in the core, not as a footnote, and it is openly recruiting Chinese-reading contributors to help finish that work.
- It sells nothing. No ads, no affiliate links, no paywall.
In a field where everyone is selling certainty, a reference you are allowed to argue with feels genuinely new.
[link] Worth a look if you are trying to figure out what to read and who to trust in AI.
If you want a sharper, shorter version
Someone built an AI reading canon you are actually allowed to argue with.
It ranks texts, not people. It shows its evidence for every call. It is built to include the American and Chinese literature, with the Chinese section still under construction. It invites public challenges and publishes every resolution. And it sells nothing, no ads, no affiliate links.
“A canon you can check is worth more than a canon you must believe.”
[link]
One honest note for whoever shares it
The rankings are still in pilot and the Chinese section is still being built. If you want to help with the second part and you read Chinese, that is an open invitation, not a disclaimer. Saying so out loud tends to make the post better, not worse.