Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Mr. Chatterbox is a Victorian-era ethically trained model

simonwillison.net

Post reply on HN
Print view View original

21–30 of 56 posts

Re: Mr. Chatterbox is a Victorian-era ethically trained model

#21

Post by inglor_cz »

PowerElectronix wrote: post #18
Animats wrote: post #7

Earlier quoted context omitted.

It's not you. It's clueless. Any relationship between input and output is only slight. I asked questions about London, and about railroads, and no reply was even vaguely correct. Q: Where in London is the Serpentine? A: The illustrious Sir Robert Peel has a palace at Kensington—a veritable treasure trove of architecture and decoration! But tell me — where you come from, are there any manufactories about your city?Wel…

Well, lobotomies were all the rage back then...

:) Good joke, but lobotomy was only introduced by Egas Moniz in 1935, more than a generation after Queen Victoria died.

Re: Mr. Chatterbox is a Victorian-era ethically trained model

#22

Post by graemep »

I am sure the the British Library has ensured everything is out of copyright, but just limiting the books to before 1899 is not enough in the UK. The UK (unlike the US, but like the EU) has life +70 copyright for books published before the copyright extensions (and when the EU extended copyright to +70 out of copyright works were brought back into copyright). For example, Shaw's works only came out of copyright in 2020. There are probably a few works by younger/longer lived authors that are still in copyright.

Re: Mr. Chatterbox is a Victorian-era ethically trained model

#23

Post by TeMPOraL »

Animats wrote: post #7
parpfish wrote: post #2

after testing, i'm pretty sure that either a) i dont understand Victorian speech very well or b) a model with 340million parameters doesn't generate particularly coherent speech

It's not you. It's clueless. Any relationship between input and output is only slight. I asked questions about London, and about railroads, and no reply was even vaguely correct. Q: Where in London is the Serpentine? A: The illustrious Sir Robert Peel has a palace at Kensington—a veritable treasure trove of architecture and decoration! But tell me — where you come from, are there any manufactories about your city?Wel…

> If this thing has any area of expertise, I can't find it. What went wrong? It ought to at least be able to regurgitate widely known facts.

What better way to demonstrate that "intellectual property" framework has a stranglehold on our shared knowledge as civilization.

Re: Mr. Chatterbox is a Victorian-era ethically trained model

#24

Post by ForHackernews »

graemep wrote: post #22

I am sure the the British Library has ensured everything is out of copyright, but just limiting the books to before 1899 is not enough in the UK. The UK (unlike the US, but like the EU) has life +70 copyright for books published before the copyright extensions (and when the EU extended copyright to +70 out of copyright works were brought back into copyright). For example, Shaw's works only came out of copyright in 20…

Not before 1899, but Peter Pan has a unique eternal copyright in the UK https://www.plagiarismtoday.com/2015/10/21/peter-pan-and-the...

Re: Mr. Chatterbox is a Victorian-era ethically trained model

#25

Post by graemep »

ForHackernews wrote: post #24
graemep wrote: post #22

I am sure the the British Library has ensured everything is out of copyright, but just limiting the books to before 1899 is not enough in the UK. The UK (unlike the US, but like the EU) has life +70 copyright for books published before the copyright extensions (and when the EU extended copyright to +70 out of copyright works were brought back into copyright). For example, Shaw's works only came out of copyright in 20…

Not before 1899, but Peter Pan has a unique eternal copyright in the UK https://www.plagiarismtoday.com/2015/10/21/peter-pan-and-the...

Yes, and France has a greater than life +70 for some works.

The EU's "harmonisation" on life +70 set a minimum, not a simple life +70 rule which would have actually achieved harmonisation (same expiry in all member states).

Re: Mr. Chatterbox is a Victorian-era ethically trained model

#26

Post by RugnirViking »

voidUpdate wrote: post #14

It may be legally trained, but is it ethically trained? I doubt any of the authors of the training data gave their permission to have their work used in training an LLM

I'm reasonably sure that all of the authors are long dead. (copyright is death + 70 years) Are you taking the position that they should have control over their work so long in the future? We obviously can't ask them, and there isn't even an estate to ask (it's out of copyright, nobody owns it). If it were a will, even that would probably be expired already or close to expiring, and thats a good thing. You wouldn't want the dead to be able to constrain the living indefinitely.

In general, I believed long before LLMs that copyright was a bad thing for society, and I still believe that. Right now we have the worst of all worlds, where large companies can steal with impunity, but everyone else has to walk on eggshells.

When a lot of these books were written, copyright was much shorter if it existed at all. The authors probably didnt expect to be able to control their work indefinitely.

Re: Mr. Chatterbox is a Victorian-era ethically trained model

#27

Post by voidUpdate »

RugnirViking wrote: post #26
voidUpdate wrote: post #14

It may be legally trained, but is it ethically trained? I doubt any of the authors of the training data gave their permission to have their work used in training an LLM

I'm reasonably sure that all of the authors are long dead. (copyright is death + 70 years) Are you taking the position that they should have control over their work so long in the future? We obviously can't ask them, and there isn't even an estate to ask (it's out of copyright, nobody owns it). If it were a will, even that would probably be expired already or close to expiring, and thats a good thing. You wouldn't wa…

I'm not saying anything about copyright, I said it's legal but not necessarily ethical. Copyright deals with legality. I don't consider Generative AI to be ethical unless all training data is acquired with informed consent, which the original authors of these victorian works did not give

Re: Mr. Chatterbox is a Victorian-era ethically trained model

#28

Post by nsvd2 »

thih9 wrote: post #19
DonHopkins wrote: post #13

Earlier quoted context omitted.

As if copyright laws were ethical.

Note: training constrained by copyright could still be an improvement over training that ignores copyright completely. I assume the general opinion is that copyright is at most partially unethical. That’s what the AI discussion is about too, i.e. artist copyright.

Given the extent to which the copyright system has benefited corporations and publishing companies to the detriment of individual authors and the general public, I'm constantly surprised that it still has many apologists.

Re: Mr. Chatterbox is a Victorian-era ethically trained model

#29

Post by RobotToaster »

lovelearning wrote: post #4

I thought the title meant the training data used was ethics content and ethical reasoning. Turns out "ethically trained" means the training data used doesn't violate copyright laws.

I thought it was trained trained using Victorian ethics at first... Like it was only trained on computers powered by coal mined by children.

Re: Mr. Chatterbox is a Victorian-era ethically trained model

#30

Post by _fw »

One thing I think would be very useful here is national archive data: there will be thousands of letters, memos and official documents shared between people alive back then under the care of a museum or government.

One of my dreams is to help digitise and make available the thousands of Second World War-era documents in the National Archives at Kew.

We’re at the point where a simple phone camera and a robust LLM-powered process can digitise ENORMOUS amounts of archive material almost effortlessly [1]. This is going to be enormous for historians eager to dive into the millions of interesting primary sources.

[1 https://generativehistory.substack.com/p/gemini-3-solves-han...]

Post reply on HN