I think that's an overly simplistic description of how LLMs work, but I take your point. My response would be: how is a LLM trained on other artists work any different to a human artist taking inspiration from other human artists? Is an artist who creates fan art of Batman also derivative? In your argument it's a clear breach of copyright, so should we be going after anyone who has ever drawn a picture of Batman as having broken the law?
YungOnions
joined 1 year ago
You son of a bitch...
You forgot to mention 'AI slop'...
Well said. These anti AI arguments tend to be a prime example of protectionism. 'Technology is great until it comes for my area of expertise, in which case my area of expertise must be protected at all costs because my work has greater value'.
The argument being made is: "AI is currently slop but there is a reasonable expectation that it will be pushed until it is indistinguishable from human work, and therefore devaluing of human work.
Again, if the work is 'indistinguishable' then I don't see how AI art 'devalues' human work any more than the work done by another human. This just sounds like old fashioned competition, which has existed as long as art itself has.
I don't like AI because it's just another way that "corporate gonna corporate" and it never ends up working out for the mere mortals' benefit
Corporations abusing technology to the disbenefit of people is nothing new, unfortunately, and isn't unique to AI (see Email, computers, clocking in machines, monitoring software etc). That speaks to a need for better corporate oversight and better worker rights.
misinformation is already so prevalent and it's going to continue to get worse (we have seen this already--trump abuses it continually).
This is a good point, but again AI is hardly the first time technology has been used to spread lies and misinformation. This highlights a fundamental problem with our media and a need to teach better critical thinking in schools etc.
They're all valid concerns but in my opinion they suggest AI is being used as an enabler, and not that the problems in question are the sole product of it. Sadly if we stopped using anything and everything that was misused for nefarious means we'd go back to the stone age.
My job here is done! Away!
This was a really interesting article. The parallels with modern day concerns are remarkable.
AI is generally bad because it tends to steal content from human creators...
Again, this is an argument that I see a lot, that's simply not true. AI is not stealing anything. Theft is a specific legal term. If I steal your TV, I have your TV and you don't. If AI is trained on some content that content still exists. Whatever training takes place steals nothing.
...because corporations want another excuse to throw more workers on the street in favor of machines...
Your point is a valid one, but this not unique to AI and is the inevitable result of the onward march of technology. The very thing we're using to communicate right now, the Internet, is responsible for billions of job losses. That's not a valid reason to get rid of it. Instead of blaming AI for putting people out work, we should be pressuring governments to implement things like UBI to provide people with a basic living wage. That way people need not fear the impact the advance of technology will have on their ability to feed and house themselves.
There are some AI uses that are good though, such as AI voice generation to help those that can't speak to communicate with the world and not sound like a robot.
These are great examples.
Sure, but there's never a qualifier in these arguments. It's just 'hur dur AI bad' which is lazy and disingenuous.
IT'S NOT A PHASE, MUM!!
Unfortunately this isn't the case. Just look at the anti AI discourse from people like Steven Zapata or Karla Ortiz, or the discussion on platforms like Mother's Basement or Art Cafe. There are plenty of artists who absolutely believe that AI art is worthless, without merit and is coming to destroy 'real artists'.