Hacker News Reader

Comment by gene-h | Hacker News Reader

Japan has explicitly amended their copyright code to enable machine learning on copyrighted data.[0] [0]https://storialaw.jp/en/service/bigdata/bigdata-12

View on HN

> although the laws of foreign countries have provisions with the same effect as Article 47-7 of Japan’s Copyright Act, all of which limit use to development for non-commercial purposes and development by research organizations

It reads like there's a bunch of countries that have similar legislation, interesting though.

Very interesting! This would give Japan a real competitive productivity advantage if ML tools are banned or severely hampered elsewhere. It would also mean that other countries would need to ban not just scraping but also the resulting code or AI-generated media.

I'm sure chatgpt would be of much less interest worldwide if it only spoke japanese.

Doesn't follow. Why would a tool built in Japan limit itself to Japanese? ChatGPT doesn't limit itself to English.

To export you must abide to the laws of the country you're exporting to.

your novel theory of extraterritorial jurisdiction will no doubt be very interesting to trips litigators

afaik asahi v. superior court is still governing precedent in the usa though so it won't be of any interest to domestic litigators in the usa

Perhaps you can't read words?

Extraterritorial jurisdiction?

You sell stuff to country A, you comply with laws of country A. Which is why USA companies have to take GRPR into account.

You make an illegal model for country A? Can't sell it there.

You wouldn't sell the model, you would sell its outputs. I can legally buy products made by factories in Bangladesh that would violate all sorts of laws if they were operating in the U.S. It will be for international trade agreements to determine.

You mean operate in a legal grey area and shut down when the loophole is closed

So has the EU as part of the digital single market changes in 2019 (The so-called Text and Data mining exceptions)

Not really.

Well they allow mining the data but nothing is said about the copyright of the collage output.

"The collage output"? What?

your novel theory of how diffusion models work will no doubt be of great interest to deep learning researchers

requires training data: yes

occasionally outputs training data verbatim: yes

verbatim output is somehow not a copyright violation: ???

Reply on news.ycombinator.com