Microsoft Deletes Blog Telling Users To Train AI on Pirated Harry Potter Books (arstechnica.com)
- Reference: 0180832630
- News link: https://it.slashdot.org/story/26/02/20/1918241/microsoft-deletes-blog-telling-users-to-train-ai-on-pirated-harry-potter-books
- Source link: https://arstechnica.com/tech-policy/2026/02/microsoft-removes-guide-on-how-to-train-llms-on-pirated-harry-potter-books/
The blog, written in November 2024 by senior product manager Pooja Kamath, walked users through building Q&A systems and generating fan fiction using the copyrighted texts, and even included a Microsoft-branded AI image of Harry Potter. The Kaggle dataset's uploader, data scientist Shubham Maindola, told Ars Technica the public domain label was "a mistake" and deleted the dataset after the outlet reached out.
[1] https://arstechnica.com/tech-policy/2026/02/microsoft-removes-guide-on-how-to-train-llms-on-pirated-harry-potter-books/
Quidiich (Score:2)
Was the game of Quiditch designed by AI?
All that matters is which sides seeker gets the snitch.
The rest of the game and team is totally irrelevant.
Re: Quidiich (Score:3)
In the World Cup, the team the caught the Snitch still lost.
But, yeah, it's a stupidly awarded game.
Likely intentionally so, though. JK said that she designed the money system (17 Sickles to a Galleon and 29 Knuts to a Sickle, making 493 Knuts equal to one Galleon) because her sister hated the old duodecimal money system in England.
So she probably had some other reason for making Quidditch such a stupid game.
Blame it on AI trained on Harry Potter... (Score:2)
Blame it on AI trained on Harry Potter...and call it the "Slytherin Effect" J.K. Rowling's literary contribution in the 21st century. ;-)
--JoshK.
Times Change (Score:2)
Normally, I'd read this article. But, now that I know that ARS Tecnica is using AI to write its stories, I just take a pass on all things AARS Technica.
Re: (Score:2)
> Normally, I'd read this article. But, now that I know that ARS Tecnica is using AI to write its stories, I just take a pass on all things AARS Technica.
That's both reductive and premature.
We know that a writer (singular) violated policy by using AI to generate a (very important) part of a story he wrote.
This is - to me - akin to finding out that a sportsball team cheated in a game. Did they cheat more than once? Don't know yet. Are more teams cheating? Don't know yet. But the matter is under investigation. In the interim, it seems reasonable to watch sportsball. If the investigation results and consequences aren't satisfactory, then it's time to
They don't even pretend to care (Score:2)
Not even any pretence to behaving legally or ethically
By marking a dataset with a flag, how could that lead anyone to believe that Harry Potter books are in the public domain and can be freely used.
It's not about the quality of the writing - if Harry Potter is fair game then what is off limits?
I guess that being in an environment where cheating is the norm gets into the blood.
Slop in, slop out (Score:2)
Harry Potter is an example of something that is popular, but not very good. The movies were better, they cleaned up a lot of the worst parts, but the books really needed some proper editing.