maegul (he/they)

A little bit of neuroscience and a little bit of computing

  • 38 Posts
  • 406 Comments
Joined 3 years ago
cake
Cake day: January 19th, 2023

help-circle

  • I’m pretty sure a major problem with AI is that people don’t know how to talk about the actual problem.

    It’s mega-corps replacing people while “stealing” their intellectual output. It’s a reshaping of the “social contract”. It’s those making bank quickly pushing the whole world in a direction that may just be a really bad idea (like we’re no longer in control of our own world/lives at all).

    It poses the question: is all tech progress good or is it sometimes “bad” such that we actually want to design our own fate and sometimes just be comfortable with what we have and what we are?

    Does this whole post-industrial revolution world need techno-hype to be happy? Or can we collectively just touch grass and settle a bit more.






  • Interestingly, I don’t think I share this sentiment.

    I’m no fan and personally don’t use AI (I barely touched it early ChatGPT days). But people use it to do things in successful fulfilment of their initial purpose.

    I’ve seen it. Maybe I’ve seen the successes and not the failures in some cases. And I’ve certainly seen badly failed attempts to use it, but in those cases I’m happy to ascribe the failure substantially to a misapplication of the tool (which to be fair certainly invites gross misapplication).

    My point though is that I don’t think an absolutist “AI is never useful” position is persuasive any more nor absolutely accurate.

    Which, in my view, makes addressing the “rest of the situation” all the more fundamental. Indeed, I think everything g other than its efficacy was always the important part.

    Part of the problem is that ethical arguments are difficult for people and many just switch off when it comes to the common good. Which is all of course part of the problem too.

    But I think that’s gravity of the situation right now: our collective instincts may be misaligned for the moment. Our personal habits vulnerable from our prior corruptions. And our societal architectures already mutated, perhaps beyond repair, and therefore ill equipped for this.

    Doomy, yes, but you’ve got to fight the fight you’re in, not the one you’d wish you’d won.

    Another way I could put this counter, is that I feel like so much of what’s bad about AI was bad before AI, and that society from 2005-2020 badly mishandled technology. Whether AI “works” or not doesn’t matter. So long as it can fit into the same shape and meet the same urges that tech did 2005-2020, it will be adopted. But if the consequences of its adoption are graver than what came before, then the whole stack of that history needs to be addressed.




  • Yea this vibes.

    For me, it felt like coding was a more attractive term for people who weren’t “proper” computer science and “engineering” types who weren’t confident that they knew what “program” meant or even “algorithm”, as they were working things as they went.

    I’d guess that as computing involved more and more people with this non-standard background, coding became preferred. I certainly encountered people uncomfortable with my casual use of “algorithm” because it triggered their imposter syndrome, and my pointing out that they write algorithms with the code the write all the time certainly didn’t help.









  • Generally, IMO, everything wrong with AI has been all the stuff other than the AI itself.

    The Capitalist urge to eat and digest the world, as well as its herd-hype mentality.

    But also the strong willingness many have had to just accept an information overlord as though it’s a religious oracle or something. All without any critical consideration of what’s happening. I blame our education systems for stagnating at some point in the past few decades — which, along with an unmitigated embrace of big corp capitalism, left us wholly unprepared for big tech’s consumption of society.

    There’s also what I’d call “the slavery urge” at play I think. At some point, an AGI will probably be conscious. But everyone is clearly so ready to turn it into our work slaves. All while pretending its output belongs to them because they “prompted it”.

    Then there’s the whole attention span being eaten thing, and quick always being ordered over good amongst an ever growing pile of increasingly shitty things.