scrap one's content

[milan2003_07]

Hello everyone,

I'm reading an article in the New York Times (electronically) and have come across the following paragraph (the article discusses the Internet, which is about to become much worse):

(here is the link at the article online: https://www.nytimes.com/2023/09/23/opinion/ai-internet-lawsuit.html)

Already, artists are deleting their work from X, formerly known as Twitter, after the company said it would be using data from its platform to train its A.I. Hollywood writers and actors are on strike partly because they want to ensure their work is not fed into A.I. systems that companies could try to replace them with. News outlets including The New York Times and CNN have added files to their website to help prevent A.I. chatbots from scraping their content.

I'm wondering about the meaning of "scrap someone's content":

Does "to scrap" mean "to steal"?

 

[jutfrank]

It's 'scrape', not 'scrap'.

It's referring to the process of extracting a text (or parts of a text), for the purposes of training LLMs. I don't really understand precisely how this works, but you don't need to to understand the point being made. As far as the writers are concerned, it's stealing of sorts.

 

[milan2003_07]

I've checked the dictionary for "scrape" and found the closest meaning to the context above. I think "to scrape someone's content" means "to steal/take something violently and maybe in an uncontrolled way" (I've figured out this definition myself based on what I've read in the dictionary).

In the situation above I think that some measures have been taken to prevent the content from being stolen by chatbots. Is it correct?

 

[emsr2d2]

In the situation above I think that some measures have been taken to prevent the content from being stolen by chatbots. Is it correct?

Basically, yes.

 

[jutfrank]

I've checked the dictionary for "scrape" and found the closest meaning to the context above. I think "to scrape someone's content" means "to steal/take something violently and maybe in an uncontrolled way" (I've figured out this definition myself based on what I've read in the dictionary).

No, there's nothing violent about it. How can an automated process be violent? And what do you mean by 'uncontrolled'?

In the situation above I think that some measures have been taken to prevent the content from being stolen by chatbots. Is it correct?

Yes. News outlets have added files to their websites.

 

[milan2003_07]

No, there's nothing violent about it. How can an automated process be violent? And what do you mean by 'uncontrolled'?

I made such a conclusion because "to scrape" also means "to remove something from a surface especially by pulling a sharp object over the surface". In the latter case we do it in quite a violent way without caring too much about the surface condition. "Uncontrolled" (maybe is should be "uncontrollable" instead) is probably incorrect here; the intended meaning was that we can make abrupt movements to reach the goal - to remove something from the surface.

 

[jutfrank]

I made such a conclusion because "to scrape" also means "to remove something from a surface especially by pulling a sharp object over the surface".

That was a very poor conclusion, don't you think? Given that we're talking about automated computer processes.

When trying to understand the use of any item of vocabulary, think about the context.

 

[milan2003_07]

That was a very poor conclusion, don't you think? Given that we're talking about automated computer processes.

When trying to understand the use of any item of vocabulary, think about the context.

Yes, I realize that this conclusion is incorrect. Actually, I first didn't manage to find the appropriate translation of the word "to scrape". Then I looked the word up in my dictionary (I've been using Collins Cobuild mainly for a long time and it proves very efficient) and found several different translations. I know that context is king and though I was argueing about it some time ago, now I realize that the context is inevitable in order to get the right answer or deduce the meaning of a word myself.

I've paid attention that you've mentioned "an automated computer process" in two of your posts. Are you saying that "to scrape one's content" is an automated process where a computer is involved?

 

[jutfrank]

Web scraping - Wikipedia

en.wikipedia.org

 

[milan2003_07]

Web scraping - Wikipedia

en.wikipedia.org

Very informative, indeed! I've studied some of the similar technologies before, but it seems to me I've come across "web scraping" for the first time.

 

[jutfrank]

So how would you say that in Russian?

 

[milan2003_07]

So how would you say that in Russian?

I've found the transliteration on the Internet in wikipedia:

https://ru.wikipedia.org/wiki/Веб-скрейпинг

Probably, it's "веб-скрапинг", but I've never worked with it before as my specialization is different. The explanation given by the Russian definition of "web scraping" is very similar to what we've been discussing.