3 ms·

Google and YouTube are trying to have it both ways with AI and copyright

76 points by isaacfrond 3y ago

frumper 3y ago
What a messy world we've built
mimd 3y ago
I'll bite verge/vergits?. The article could have a better title, as the article clearly states Google is more in a dispute with one of its suppliers , and less of believing UMG's position. It's less hypocritical as implied by the title. The articles could start mentioning the CFAA, and how US courts have interpreted it. It's kinda important to how the web scraping questions will go in court. I'd like an article about world wide web scraping legalities. Also, verge should mention in these articles that they have said no to gpt bot. https://www.theverge.com/robots.txt https://www.theverge.com/robots.txt. Not that it really matters in figuring out vox's positions, but they like to report on conflicts of interest.
- AshamedCaptain 3y ago
  I really really don't see the "conflict of interest" in blacklisting ChatGPT. What are going to accuse them of, actually practicing the opinion they seem to be publishing ? What's next, if you publish an article of politician A, you have to disclaim that you didn't vote for him in the last election ? "Conflict of interest" is normally reserved for when you are politician B, or paid by B... e.g., OpenAI complaining about Google or the like.
  - mimd 3y ago
    It's a little rap on the hands, a tease, considering they love highlighting conflicts of interest when it doesn't concern them. COI's don't have much legal binding on US news reporting, unless it comes down from courts. They're more of courtesy and an attempt to get some of that legal veneer. But it is often done to build trust with the reader and not underhandedly trying to cast yourself as impartial third party who can be trusted to give a best effort picture of the situation. Ie, vox media is currently taking steps against chatgpt potentially leading to a lawsuit, why should you trust their reporting to be fair or even thorough on the subject? It's not a question of if they are right on their lawsuit, but if they giving you reasonable service. Now I suspect your fine with that service, but I like a bit more impartiality and thoroughness and the imbalanced use of COI's smells. COI's can get very draconian in other sectors with legal binding s. It does nearly reach who you vote for. And practicing your "Opinion" doesn't give you a legal basis to breach those bounds, and say insider trade off a company.
AshamedCaptain 3y ago
> A website could block Google’s crawlers in its robots.txt file — OpenAI, fresh from scraping every website in the world to build ChatGPT, just allowed its crawler to be blocked in this way — but blocking Google’s crawlers means deindexing your site from search, which is, bluntly, suicidal. And someone still doesn't think Google is constantly abusing their monopoly position ? This is basically textbook example, and it's not the first time they do something like this.
nuc1e0n 3y ago
For the sake of clarity, I think Google shouldn't be using the same bot to crawl the web for search results as well as for training LLMs. Has it been legally challenged by the New York Times that that is what Google appears to be doing right now? Bard just isn't the same thing as Google Search.
- nadermx 3y ago
  How so? Both cache some sort of data and regurgitate back something from that data at behalf of the users request.
  - WorldMaker 3y ago
    One is built to provide that data back in the context of search links to the source and the other attempts to synthesize the data with other subject matter and may not be accurate at all in linking back to original sources. The output and intended use cases are quite different even if the sausage processes at times seem similar.
  - nuc1e0n 3y ago
    yeah, what WorldMaker said

76 points by isaacfrond 3y ago

What a messy world we've built

mimd 3y ago

I'll bite verge/vergits?. The article could have a better title, as the article clearly states Google is more in a dispute with one of its suppliers , and less of believing UMG's position. It's less hypocritical as implied by the title. The articles could start mentioning the CFAA, and how US courts have interpreted it. It's kinda important to how the web scraping questions will go in court. I'd like an article about world wide web scraping legalities. Also, verge should mention in these articles that they have said no to gpt bot. https://www.theverge.com/robots.txt https://www.theverge.com/robots.txt. Not that it really matters in figuring out vox's positions, but they like to report on conflicts of interest.

AshamedCaptain 3y ago
I really really don't see the "conflict of interest" in blacklisting ChatGPT. What are going to accuse them of, actually practicing the opinion they seem to be publishing ? What's next, if you publish an article of politician A, you have to disclaim that you didn't vote for him in the last election ? "Conflict of interest" is normally reserved for when you are politician B, or paid by B... e.g., OpenAI complaining about Google or the like.
- mimd 3y ago
  It's a little rap on the hands, a tease, considering they love highlighting conflicts of interest when it doesn't concern them. COI's don't have much legal binding on US news reporting, unless it comes down from courts. They're more of courtesy and an attempt to get some of that legal veneer. But it is often done to build trust with the reader and not underhandedly trying to cast yourself as impartial third party who can be trusted to give a best effort picture of the situation. Ie, vox media is currently taking steps against chatgpt potentially leading to a lawsuit, why should you trust their reporting to be fair or even thorough on the subject? It's not a question of if they are right on their lawsuit, but if they giving you reasonable service. Now I suspect your fine with that service, but I like a bit more impartiality and thoroughness and the imbalanced use of COI's smells. COI's can get very draconian in other sectors with legal binding s. It does nearly reach who you vote for. And practicing your "Opinion" doesn't give you a legal basis to breach those bounds, and say insider trade off a company.

AshamedCaptain 3y ago

> A website could block Google’s crawlers in its robots.txt file — OpenAI, fresh from scraping every website in the world to build ChatGPT, just allowed its crawler to be blocked in this way — but blocking Google’s crawlers means deindexing your site from search, which is, bluntly, suicidal. And someone still doesn't think Google is constantly abusing their monopoly position ? This is basically textbook example, and it's not the first time they do something like this.

nuc1e0n 3y ago

For the sake of clarity, I think Google shouldn't be using the same bot to crawl the web for search results as well as for training LLMs. Has it been legally challenged by the New York Times that that is what Google appears to be doing right now? Bard just isn't the same thing as Google Search.

nadermx 3y ago
How so? Both cache some sort of data and regurgitate back something from that data at behalf of the users request.
- WorldMaker 3y ago
  One is built to provide that data back in the context of search links to the source and the other attempts to synthesize the data with other subject matter and may not be accurate at all in linking back to original sources. The output and intended use cases are quite different even if the sausage processes at times seem similar.
- nuc1e0n 3y ago
  yeah, what WorldMaker said