Every novel that has ever been published is sitting inside ChatGPT (opens in new tab)

I truly wish to see Google/Alphabet be absolutely annihilated by lawsuits. Bad enough that YouTube was built on pirated material, and still is to this day, but now this?

Every single penny they’ve ever generated should be awarded to the authors they stole from, and Alphabet should be bankrupt into oblivion. The sheer number of people involved in this mass piracy event means it’s fully systemic. Shut them down!

> Over 190,000 copyrighted books obtained from pirated websites.

While that’s a lot, the estimated number of just English published novels is roughly an order of magnitude above that [0], so “every novel that has ever been published” isn’t anywhere near correct, in all likelihood.

[0] https://litlab.stanford.edu/how-many-novels-have-been-publis...

This tweet is misleading (shocker, I know; Twitter and misleading ragebait—who could have guessed?).

It claims that "Up to 90%" (accurate, or at least plausible—but unsurprising) of "Every book you have ever read" (just untrue) is "sitting inside ChatGPT right now".

Meanwhile, 100% of the books that have been scanned into Google Books's scanned books collection are sitting "inside" Google Books's scanned books collection. And 100% of the web pages that Google Search has crawled and indexed are sitting "inside" Google Search's index of the pages it has crawled and indexed.

Is it just me or is this full of LLMisms? Threes, "not X, Y", etc.