Post B4Eu5VYXNSACZGibbs by skyfaller@jawns.club
 (DIR) More posts by skyfaller@jawns.club
 (DIR) Post #B4EtxBQYKdB7p50wWe by xgranade@wandering.shop
       0 likes, 0 repeats
       
       I feel like if I were on the other side, and I wanted to advance AI, I'd focus a lot less on rhetorical tricks and a lot more on taking the criticisms seriously and making sure that what I'm advocating is actually ethical.I don't mean local models here, either. The problems with AI are structural, as much social as technical, and aren't solved by shifting the goalposts.
       
 (DIR) Post #B4EtxBdJZBO2Sez8Pw by xgranade@wandering.shop
       0 likes, 1 repeats
       
       If I'm wrong, if everyone else setting off every alarm bell they can is wrong, do the fucking work and prove us wrong. Go build an actual ethical basis for AI that isn't just an "alignment" sleight of hand. Go pierce corporate secrecy and get real numbers. Go figure out how to keep fascists from using AI to demolish truth. Go figure out how to deal with devastating effects on common culture. Go build labor rights and prevent AI from being an automated scab.
       
 (DIR) Post #B4EtxG435j9GCjWXdw by xgranade@wandering.shop
       0 likes, 0 repeats
       
       The problem isn't convincing your audience that I'm wrong, it's *proving to everyone you can, yourself included* that I'm wrong.Whataboutism, Gish galloping, ignoring burden of proof, relying on biased estimates, assuming inevitability, appealing to distorted understandings of "harm reduction," conflating sci-fi tropes with marketing terminology, all of that is just rhetoric. It doesn't address the substance of the problem.
       
 (DIR) Post #B4EtxHQm0iv6RVAChc by xgranade@wandering.shop
       0 likes, 0 repeats
       
       Maybe put differently, if you are advancing AI as a cause: why? What are you trying to accomplish, and how will that make the world a better place? How can that be undermined if the AI Luddite view¹ is correct? What can you do to make sure your advocacy reflects your intent?This is the work of any kind of activism or advocacy! AI isn't specially exempt here!___¹in the labor rights sense, not the pejorative sense
       
 (DIR) Post #B4EtxIuaVMMZ2G7WoS by xgranade@wandering.shop
       0 likes, 0 repeats
       
       As I keep coming back to, I'm one angry loud woman on Satan's own internet. Convincing me of something doesn't change shit.I raise the issues I do because I think they're factually correct and morally relevant — if I'm wrong *and you believe in all good faith and due diligence that I am wrong*, then yeah, sure, ignore me. But trying to *convince* me with rhetoric isn't accomplishing anything.
       
 (DIR) Post #B4Eu1ra9XjAdipjhke by zombiecide@polyglot.city
       0 likes, 1 repeats
       
       @xgranade I took model building and simulation this semester. In the introduction we learnt that a model needs to have a specific use case to be implementable.Following this, the specific use case for genAI is: "produce output that can fool a layperson". By design. Nothing else can come out of it.And the use case for other forms of predictice analytics that compile data on people without their knowledge or consent is to  be able do things they would never consent to.
       
 (DIR) Post #B4Eu5VYXNSACZGibbs by skyfaller@jawns.club
       0 likes, 1 repeats
       
       @xgranade AI boosters will never actually do this, tho they might claim to. All they want are excuses to ignore damage from AI, so they can do what they want to without criticism or pushback. It's not good enough to have all the money in the world on their side, they don't even want to see dissent.Instead they will tell you how it's not as harmful as you think, or it's less damaging than other things people accept, or the harms will be fixed soon, or it's worth it to build an enslaved god.