There's a temptation to be "nuanced" or to hold "complex" views about AI.
-
There's a temptation to be "nuanced" or to hold "complex" views about AI.
I ask that you please reject that temptation and recognize that full and complete abstinence from AI is a complex view that considers the tradeoffs, compromises with the relative extremes of direct action, and that comes from a comprehensive understanding of what AI is.
@xgranade This guy's view is very sobering: https://medium.com/@igisho/the-ai-race-we-cant-win-or-why-we-re-treating-stagnation-with-poison-ce0cc9024e2a
-
To wit, AI products such as LLMs:
• Are primarily developed under eugenicist philosophies, and primarily financially benefit fascists.
• Are trained on uncompensated and nonconsentual labor.
• Require unconscionable and unsustainable levels of energy and water waste.
• Are used to undermine labor rights, such as the power of artists to say "no" to unconscionable demands.
• Necessarily consolidate and centralize power.
• Present unacceptable risks due to disinformation and other errors.@xgranade
Plausible Whatever Generators also ideologically support fascists by enabling the mass production and popularization of imagery that is wholly detached from reality. there's a bit more nuance to the pile. -
J jwcph@helvede.net shared this topic
-
@xgranade Ironically my telling people "Beyond all my moral objections, I'm simply disinterested in LLMs (or 'AI' as you're calling it) to such an extent that I'd rather leave this field I care deeply about" appears to carry a level of nuance that simply doesn't seem to compute to many of them!
So much so that I needed to turn my passion into a talk...
-
There's a temptation to be "nuanced" or to hold "complex" views about AI.
I ask that you please reject that temptation and recognize that full and complete abstinence from AI is a complex view that considers the tradeoffs, compromises with the relative extremes of direct action, and that comes from a comprehensive understanding of what AI is.
@xgranade exactly! None of this "I'm an AI skeptic" or "AI optimist" nonsense. Commit to full rejection. Don't take the AI industry's linkedin bait.
-
(Similarly, I might well be criticized for discussing things in relatively idealistic terms, and ignoring how one might be *coerced* into using or promoting AI. I will defend my discussion on idealistic terms, as "AI is bad, don't use it because of this list of reasons" is already a nuanced and complex statement about what one should ideally do. Adding the other layer of constrained choices under capitalistic coercion is another necessary complexity, but a separate and complementary discussion.)
@xgranade why the demand for “nuance”? When something is just hands down bad, period, i refuse to be forced to carefully lay out all the “complex” reasons that everyone already has seen. Is there a nuanced critique of napalm? No. None is needed.
-
(I say possibly because it is often fair to summarize a nuanced and complex position in a perhaps overly simplistic way, due to the exigencies of short-form discussions. A terse view might be ill-considered, or it might be a casual summary of a complex and detailed consideration. I don't suggest holding good faith or grace about that ambiguity, given how many bad faith actors try to fuck up AI discourse, but I only suggest that *if* someone acts in good faith, they may be summarizing.)
@xgranade@wandering.shop
Counterpoints:
- relying on AI in the workplace has kept me employed
- using AI as a supplemental tool to my journal, notes, mood trackers, health trackers, todo lists has helped me actually do things (not do things better, but do things in the first place)
If you have an alternative for either of these functions that is as affordable, I'm very open to your recommendations.
Which is to say, your position isn't nuanced or complex, unless you account for all sides of this debate:
- the creatives whose craft/ art was used for training AI,
- the companies doing the training,
- the end users who are effectively coerced into using AI,
- the harm-reduction activists such as yourself,
- the multi-national political powers involved,
- the military and intelligence factors of the technology,
- etc. etc.
Currently you only account for one side. You may say "the other sides are wrong," but being stuck in a job where AI is a requirement, not an option... show me an alternative that won't screw me over. -
That is not a simple position to hold, but precisely none of those complexities point to embracing the relatively extreme position of AI adoption, let alone AI boosterism.
Nuanced and complex positions are not necessarily those that coincide with popular discourse. You are not necessarily a more careful thinker for hedging your own internal cognition towards media narratives.
That's a really good point.
I'd like to add that you are not self-disciplined for suppressing your conscience.
-
There's a temptation to be "nuanced" or to hold "complex" views about AI.
I ask that you please reject that temptation and recognize that full and complete abstinence from AI is a complex view that considers the tradeoffs, compromises with the relative extremes of direct action, and that comes from a comprehensive understanding of what AI is.
@xgranade So very true. See also: https://svpow.com/2014/03/20/a-nuanced-position-isnt-always-correct/
-
There's a temptation to be "nuanced" or to hold "complex" views about AI.
I ask that you please reject that temptation and recognize that full and complete abstinence from AI is a complex view that considers the tradeoffs, compromises with the relative extremes of direct action, and that comes from a comprehensive understanding of what AI is.
I'd rather no AI, but that horse has left the barn - AI is not going away; so not wise IMO to not make use of its positive aspects. There are many aspects to AI - not just slop.
H (Human) I is failing miserably (disinformation, fascism, climate change, etc.). AI carries with it existential risks, but also enormous potential for good: evidence-based information at everyone's fingertips, medicine, personalized education.
Regulating AI is essential.
-
(Similarly, I might well be criticized for discussing things in relatively idealistic terms, and ignoring how one might be *coerced* into using or promoting AI. I will defend my discussion on idealistic terms, as "AI is bad, don't use it because of this list of reasons" is already a nuanced and complex statement about what one should ideally do. Adding the other layer of constrained choices under capitalistic coercion is another necessary complexity, but a separate and complementary discussion.)
@xgranade Honestly, it just sounds like the way vegans talk about veganism. "As far as is possible and practicable." is a phrase that is often used.
(Am vegan, this is not criticism, the opposite actually)
-
(Similarly, I might well be criticized for discussing things in relatively idealistic terms, and ignoring how one might be *coerced* into using or promoting AI. I will defend my discussion on idealistic terms, as "AI is bad, don't use it because of this list of reasons" is already a nuanced and complex statement about what one should ideally do. Adding the other layer of constrained choices under capitalistic coercion is another necessary complexity, but a separate and complementary discussion.)
@xgranade IMO the only " Nuance" I personally would suggest in this is that AI /= GenAI. Like GPS is not NEARLY as bad as ChatGPT and all that with LLMs. AI can have it's uses, but LLMs are practically being sold as snake oil if that makes sense.
Edit: I should add this goes for Pro-AI people saying stuff like "AI is here to stay", "Everyone uses AI" and stuff like that.
-
There's a temptation to be "nuanced" or to hold "complex" views about AI.
I ask that you please reject that temptation and recognize that full and complete abstinence from AI is a complex view that considers the tradeoffs, compromises with the relative extremes of direct action, and that comes from a comprehensive understanding of what AI is.
Not only "not use" but actively resist AI, actively poison scraping theft, actively disable on websites, actively hide with uBlock.
_______
AI is soldiers. Soldiers of an occupying army taking our jobs, our energy, and our water.
______
AI is shyte.
CLOSE DATA CENTRES.
Oppose data centers like ICE concentration camps. -
There's a temptation to be "nuanced" or to hold "complex" views about AI.
I ask that you please reject that temptation and recognize that full and complete abstinence from AI is a complex view that considers the tradeoffs, compromises with the relative extremes of direct action, and that comes from a comprehensive understanding of what AI is.
@xgranade similar to “No is a complete sentence”
-
There's a temptation to be "nuanced" or to hold "complex" views about AI.
I ask that you please reject that temptation and recognize that full and complete abstinence from AI is a complex view that considers the tradeoffs, compromises with the relative extremes of direct action, and that comes from a comprehensive understanding of what AI is.
I fully understand what AI is and this is why in my mind there are only two tasks where it's a genuine improvement over anything that came before: natural language processing (the kind voice assistants do) and text summarization. But, anything that relies on its internal "knowledge", which is most things that AI is being pushed for right now, is laughably crap and will always be.
Another useful trait is when you see an article with a generated illustration, you instantly know that it's not worth your time.
-
There's a temptation to be "nuanced" or to hold "complex" views about AI.
I ask that you please reject that temptation and recognize that full and complete abstinence from AI is a complex view that considers the tradeoffs, compromises with the relative extremes of direct action, and that comes from a comprehensive understanding of what AI is.
-
There's a temptation to be "nuanced" or to hold "complex" views about AI.
I ask that you please reject that temptation and recognize that full and complete abstinence from AI is a complex view that considers the tradeoffs, compromises with the relative extremes of direct action, and that comes from a comprehensive understanding of what AI is.
@xgranade there is very little nuance about LLMs other than the nuances of how bad they are
-
S sebastian@social.itu.dk shared this topic
