Live data from Hacker News

The shady world of Brave selling copyrighted data for AI training

stackdiary.com

1–10 of 127 posts

Re: The shady world of Brave selling copyrighted data for AI training

#4
post #2

How long until IP works its way onto ai training data or ais themselves? Ie that for some specific instance, the training is intentionally wrong, so as to check and prove that there has been a breach of IP.

Do you mean like trap streets? Seems like a good idea for model makers

https://en.wikipedia.org/wiki/Trap_street

Re: The shady world of Brave selling copyrighted data for AI training

#5
I firmly believe that corps like these don't deserve the benefit of the doubt. Google, Brave and really anyone big enough to allow themselves to do bad things and get away with it must adhere to a standard where they proactively show their stuff doesn't have malicious intents.

Re: The shady world of Brave selling copyrighted data for AI training

#6
post #2

How long until IP works its way onto ai training data or ais themselves? Ie that for some specific instance, the training is intentionally wrong, so as to check and prove that there has been a breach of IP.

While not intentionally wrong, Van Halen's brown M&M's rider comes to mind as an example of a similar measure.

Re: The shady world of Brave selling copyrighted data for AI training

#7
post #5

I firmly believe that corps like these don't deserve the benefit of the doubt. Google, Brave and really anyone big enough to allow themselves to do bad things and get away with it must adhere to a standard where they proactively show their stuff doesn't have malicious intents.

As always, if the product is free, you are the product...

Re: The shady world of Brave selling copyrighted data for AI training

#8
post #2

How long until IP works its way onto ai training data or ais themselves? Ie that for some specific instance, the training is intentionally wrong, so as to check and prove that there has been a breach of IP.

Do you mean like trap streets? Seems like a good idea for model makers https://en.wikipedia.org/wiki/Trap_street

Yeah, something like that may be already happening and various actors building their cases as we speak.

Re: The shady world of Brave selling copyrighted data for AI training

#9

I think this title is overstated. It seems like Brave is trying to do the right thing here vs other companies that don't even make the attempt. (Also, crawling as a service has been a thing for a while.)

> It seems like Brave is trying to do the right thing here vs other companies that don't even make the attempt

I feel like I'm missing something. What the article claims they're doing is:

1. Misrepresenting what rights they have, and selling access to those rights.

2. Stealth-crawling the web, hiding from the webmasters just how much Brave is crawling their site, and making it impossible to block just their crawler.

How is either of these the right thing? I mean, for somebody besides Brave. What "attempt" are they making that other companies aren't?

Re: The shady world of Brave selling copyrighted data for AI training

#10
post #5

I firmly believe that corps like these don't deserve the benefit of the doubt. Google, Brave and really anyone big enough to allow themselves to do bad things and get away with it must adhere to a standard where they proactively show their stuff doesn't have malicious intents.

As always, if the product is free, you are the product...

“Although the saying tells us “If it’s free, then you are the product,” that is also incorrect. We are the sources of surveillance capitalism’s crucial surplus: the objects of a technologically advanced and increasingly inescapable raw-material-extraction operation. Surveillance capitalism’s actual customers are the enterprises that trade in its markets for future behavior.”

Excerpt From The Age of Surveillance Capitalism Shoshana Zuboff

Post reply on HN