Jeff Atwood@nerdpr0f @nopatience @cR0w @cloudskater sure, every new tool has downsides and should never be viewed as a panacea, perfect, or works for every possible situation. TV melted many, many brains.. the whole fox propaganda machine..
cloudskater@codinghorror @nerdpr0f @nopatience @cR0w except TV as a medium is not stupifying, conservative propaganda is. Meanwhile, Generative AI, no matter what is created with it, is harmful from the offset. There's no positive use of it.
sbysb@cloudskater @codinghorror @nerdpr0f @nopatience @cR0w
I think this ends up being a reductive view that ends up pushing the people who want to do good with it away from using generative AI as a tool, leaving only the worst people you know to benefit from it's uses.
Someone using generative AI to build a community food drive website is in direct opposition to the ghouls using it to write algorithms to maximize food profits at the large grocers. And only one of those people is going to be convinced the tool itself is harmful
@sbysb @codinghorror @nerdpr0f @nopatience @cR0w
Your last sentence is correct: you will not convince evil people that they should not use the evil thing because it is evil, when they themselves created the evil thing with intentions to be evil.
However, using an inherently corrupt tool for an unambiguously good cause does not make usage of the tool any better in and of itself. Assuming they are a decent person, they will hopefully abandon the technology when informed of it's harmful nature.
As a heroic man-bat once said...
@cloudskater @sbysb @nerdpr0f @nopatience @cR0w nuclear weapons and guns are a different thing, physical danger -- and I completely agree -- the better reference is perhaps "Ghost Dance" from Alan Moore's (epic) Swamp Thing run.
@codinghorror @sbysb @nerdpr0f @nopatience @cR0w I hate to disappoint but I don't actually know much about comic history, I just used that image because its a common meme thrown at "positive" uses of "gen" AI lol
@cloudskater @sbysb @nerdpr0f @nopatience @cR0w knowing history is useful so we can learn from it and avoid prior mistakes eg https://rgmii.org/history-of-gmi/
Christoffer S.@codinghorror Reminded me of this reasonably recent experiment in Finland:
