.png "Cohost PLUS!")
There's a very weird argument I've seen for AI art bullshit from certain people that I think should know better which goes like, "Well, the people pushing it hard are wrong and are using it unethically, but I see great use cases for it when used ethically! I'm simply a curious bystander waiting for it to become acceptable to use!" and it drives me insane. The world in which "AI" (machine learning, there is no intelligence or originality here) could be used ethically does not exist. This shit is constantly being pushed as being a magic button that spits out the thing you want when you push it. The pitch that it empowers creatives to be able to make things faster is a very transparent cover for the actual pitch: you, the person who wants to make money, wants to make a thing to sell without having to deal with those pesky creatives that want to be paid for making the thing. It is an idea whose magnetic poles draw in suckers and assholes who want to make money and don't care how.
Also, the idea that AI will empower creatives by automating things has already been covered by a whole bunch of things for decades now. It's stuff like content-aware fill, SpeedTree, terrain generators, etc.
This shit makes me cranky. Makes me feel like I'm getting snobbier and more of a gatekeeper. If you want to create a thing and don't want to pay somebody to create, then fucking learn how to do it yourself. See how long it takes to be able to make the thing you want, I'll wait. Maybe along the way you'll actually gain a human soul.
the "AI" goldrush reminds me 100% of the crypto/nft goldrush because it's functioning in exactly the same way: a new technology is created, the technology has no clear usecase that would make it more useful than a previous technology, but usecases continue to be pushed by those in the power to proliferate the technology because if you can simply prove that there are enough possible usecases, then it would be stupid for someone to oppose the technology. There's simply too much potential.
Never mind, of course, that the technology is not actually usable in any of the use cases. Cryptocurrency/blockchain is an unbelievably slow, clunky, wasteful method of storing wealth or transferring wealth, and absolutely godawful for everyday purchases. NFTs offer zero advantages over a centralized ledger for trust/verification and open up thousands of possible attack vectors (many of which are just holdovers from those same attack vectors in the blockchain technologies they are built on).
ChatGPT/similar AI "services" have gotten extremely good at being believable conversation targets that actually suck ass whenever you ask them to provide simple, verifiable information. As endpoint customer service tool, they are abysmal.
Imagine having an employee who is unaccountable, prone to trailing off, and occasionally (& randomly) terrifyingly angry or hostile to its conversation partner. This is not a viable usecase for the technology.
And yet, they soldier onward, since the goal is not, actually, a useful tool -- the goal is a tool that can be sold to starry-eyed audiences that can either a) make someone money and/or b) take costs (that is to say, paychecks) out of the equation for a business.
There is no "ethical usage of AI" as it is being currently touted. It is a solution that its creators are desperately trying to buy a problem for.
