4 ms·

> alleges that Carlin’s copyrighted materials and likeness were used without permission or appropriate licenses It sounds like Carlin's estate doesn't understa

> alleges that Carlin’s copyrighted materials and likeness were used without permission or appropriate licenses It sounds like Carlin's estate doesn't understand how AI works. If it's illegal to use Carlin's material in this way, then all the popular AI models must be illegal as well lol

yladiz 3y ago
This exact question is currently being litigated, and using copyrighted material when training an LLM or AI model without permission could very well be illegal.
- MassiveQuasar 3y ago
  Which line of reasoning allows natural intelligence do the the same thing artificial intelligence isn't allowed to do.
  - HeatrayEnjoyer 3y ago
    IANAL but just spitballing: - The qualitative differences. You can't run an unbounded # of simultaneous and 24/7 instances of a natural human intelligence. - Machine learning models are not people and do not have rights. - Obtaining the training material itself requires massive copyright infringement. GPT-3 was even openly trained on books from a literal pirate torrent tracker.
    - dorkwood 3y ago
      > GPT-3 was even openly trained on books from a literal pirate torrent tracker What's stopping a human from reading all of the same books?
      - rectang 3y ago
        As the previous commenter indicated, scale — it’s not physically possible for a human to read that many books. Even the most dedicated readers will only read a few thousand books in their lifetime. https://www.mentalfloss.com/article/93594/heres-how-many-books-you-can-expect-read-you-die https://www.mentalfloss.com/article/93594/heres-how-many-boo... ChatGPT is trained on orders of magnitude more books than a human can consume. https://torrentfreak.com/authors-accuse-openai-of-using-pirate-sites-to-train-chatgpt-230630/ https://torrentfreak.com/authors-accuse-openai-of-using-pira... > OpenAI hasn’t disclosed the datasets that ChatGPT is trained on, but in an older paper two databases are referenced; “Books1” and “Books2”. The first one contains roughly 63,000 titles and the latter around 294,000 titles
- scheeseman486 3y ago
  Not anymore. https://www.nytimes.com/2024/01/26/arts/carlin-lawsuit-ai-podcast-copyright.html https://www.nytimes.com/2024/01/26/arts/carlin-lawsuit-ai-po... More likely the question is going to be how far parody and satire can allow for the use celebrity likeness, a much blurrier line and not anywhere near cut and dried; you only have to look at sketch comedy to see a whole lot of impressions where name, likeness and voice were all imitated for the purposes of comedy. Will Sasso himself did a bunch of that while he performed on MadTV. It also has little to with AI. The fidelity of the imitation is higher, but is that the biggest issue when the content in question makes it plainly obvious that it's an impression? No one was fooled that it was Carlin, they were fooled that it was written by an AI.
chaosmonger7 3y ago
This comment makes no sense. I can't put out an album of my shitty music and market it as a new Michael Jackson album. Putting "AI" in front of "Michael" doesn't change anything.
- scheeseman486 3y ago
  If the content of the music is filled with commentary and jokes about the use of AI to create facsimiles of celebrities and comedians and the exploitation of the technology for the purposes of mass commercialization by billionaires, you might be able to. After all, you can open a store selling shitty coffee and market it as Dumb Starbucks.

yladiz 3y ago

This exact question is currently being litigated, and using copyrighted material when training an LLM or AI model without permission could very well be illegal.

MassiveQuasar 3y ago
Which line of reasoning allows natural intelligence do the the same thing artificial intelligence isn't allowed to do.
- HeatrayEnjoyer 3y ago
  IANAL but just spitballing: - The qualitative differences. You can't run an unbounded # of simultaneous and 24/7 instances of a natural human intelligence. - Machine learning models are not people and do not have rights. - Obtaining the training material itself requires massive copyright infringement. GPT-3 was even openly trained on books from a literal pirate torrent tracker.
  - dorkwood 3y ago
    > GPT-3 was even openly trained on books from a literal pirate torrent tracker What's stopping a human from reading all of the same books?
    - rectang 3y ago
      As the previous commenter indicated, scale — it’s not physically possible for a human to read that many books. Even the most dedicated readers will only read a few thousand books in their lifetime. https://www.mentalfloss.com/article/93594/heres-how-many-books-you-can-expect-read-you-die https://www.mentalfloss.com/article/93594/heres-how-many-boo... ChatGPT is trained on orders of magnitude more books than a human can consume. https://torrentfreak.com/authors-accuse-openai-of-using-pirate-sites-to-train-chatgpt-230630/ https://torrentfreak.com/authors-accuse-openai-of-using-pira... > OpenAI hasn’t disclosed the datasets that ChatGPT is trained on, but in an older paper two databases are referenced; “Books1” and “Books2”. The first one contains roughly 63,000 titles and the latter around 294,000 titles
scheeseman486 3y ago
Not anymore. https://www.nytimes.com/2024/01/26/arts/carlin-lawsuit-ai-podcast-copyright.html https://www.nytimes.com/2024/01/26/arts/carlin-lawsuit-ai-po... More likely the question is going to be how far parody and satire can allow for the use celebrity likeness, a much blurrier line and not anywhere near cut and dried; you only have to look at sketch comedy to see a whole lot of impressions where name, likeness and voice were all imitated for the purposes of comedy. Will Sasso himself did a bunch of that while he performed on MadTV. It also has little to with AI. The fidelity of the imitation is higher, but is that the biggest issue when the content in question makes it plainly obvious that it's an impression? No one was fooled that it was Carlin, they were fooled that it was written by an AI.

chaosmonger7 3y ago

This comment makes no sense. I can't put out an album of my shitty music and market it as a new Michael Jackson album. Putting "AI" in front of "Michael" doesn't change anything.

scheeseman486 3y ago
If the content of the music is filled with commentary and jokes about the use of AI to create facsimiles of celebrities and comedians and the exploitation of the technology for the purposes of mass commercialization by billionaires, you might be able to. After all, you can open a store selling shitty coffee and market it as Dumb Starbucks.