Sarah Silverman suing OpenAI. So are other authors.

Mindy D · ‎05-23-2010

AI models that can write get the background needed to do so by scanning thousands of writings on the internet, including Twitter and by scanning and taking data (scraping) thousands of other types of literature including books, news reports, essays and plays. This is how the models acquire language. This is why these types of AI are called language learning models.

Sarah Silverman wrote a memoir, "The Bedwetter." If a user of ChatGBT, the AI owned by OpenAI, asks ChatGPT to write a synopsis of " The Bedwetter" it can do so. There are questions about the sources of information necessary to do this. Did the model get its information from a pirated copy of the book or did it scrape reviews of the book online?

Silverman is suing OpenAI for copyright infringement because she did not give permission for OpenAI to ingest the digital version of her book to help train its models. Silverman alleges that the copy of the work used to train the model was pirated and was used without consent and without compensation.
Silverman and other authors have a class action suit about this.

OpenAI is not the only company being sued by writers over this. Silverman and others are also engaged in another lawsuit against Meta, the parent company of Facebook, which has a different AI model.

Other authors, including Nora Roberts sent a letter to CEOs at Google, Meta, OpenAI and Microsoft which accused them of exploitive practices in the building of their chatbots, saying that the bots mimic and regurgitate their language, writing style and ideas. The letter, which was organized by The Author's Guild was signed by over 4000 writers and says that millions of copyrighted works including articles, books and poetry have provided the food for these AI systems and that they want fair compensation for their work.

Due to the use of AI companies scraping Twitter to train their AI models, Elon Musk has started limiting the viewing of a specified number of tweets per day. Musk has stated he was doing this due to the extreme number of data scrapes on Twitter.

Tinkrbl44 · ‎08-23-2010

@Mindy D

My prediction: Legally, this is going to get very messy.

Mz iMac · ‎07-26-2014

Skynet in its infancy.

If you've seen The Terminator franchise movies, you know what I mean.

"Never argue with a fool. Onlookers may not be able to tell the difference."

Goldengate8361 · ‎01-27-2014

I think ms Silverman has a strong case. I'm glad she's suing.

gardenman · ‎03-09-2010

It gets complicated. Every writer learned to write by reading. That's what AI does. It's learning to write by reading what others have written. Should writers Sarah Silverman read sue her since she learned to write from them? Can she prove AI didn't acquire her work legitimately?

Fly!!! Eagles!!! Fly!!!

Laura14 · ‎06-09-2014

As a published author, I fully support the forced legal action to protect creatives and our copyrights.

Either the copyright I earned through my hard work and dedication and paid for means something or it doesn't.

Readers inspired by writers are not turning around and making a profit in seconds off of someone else's actual months/years long efforts.

Amazon has been flooded by AI books and it makes it difficult for readers to find what they are looking for. Indie authors in particular are hard hit with even further lack of visibility.

We need rules about what's inspiration and what's flat out theft and violation of copyright. I'd also like to see mandatory labeling of AI generated books.

Mindy D · ‎05-23-2010

@Laura14 wrote:
As a published author, I fully support the forced legal action to protect creatives and our copyrights.

Either the copyright I earned through my hard work and dedication and paid for means something or it doesn't.

Readers inspired by writers are not turning around and making a profit in seconds off of someone else's actual months/years long efforts.

Amazon has been flooded by AI books and it makes it difficult for readers to find what they are looking for. Indie authors in particular are hard hit with even further lack of visibility.

We need rules about what's inspiration and what's flat out theft and violation of copyright. I'd also like to see mandatory labeling of AI generated books.

@Laura14 You are so right. I didn't know about the AI books on Amazon either. Your last two lines say exactly what's needed.

Plus, readers inspired by writers had to read from sources that were paid for by libraries, themselves or others, or were provided for free...so the authors were compensated when necessary. AI sources are not known, and Silverman and others are alleging that the sources are pirated and used without permission or compensation to the authors.

Laura14 · ‎06-09-2014

@Mindy D If humans have to cite our sources to avoid legal jeopardy, so does AI. I know this world has gone nuts, but when the software program has more rights than I do, even with a government copyright, we're in deep trouble.