1y ago

pluralistic: the alarmism around AI is just a marketing spin.

Pluralistic: The real AI fight (27 Nov 2023) – Pluralistic: Daily links from Cory Doctorow

The alarmism around AI is just a marketing spin.

As @pluralistic@mamot.fr wrote: that's "mystical nonsense about spontaneous consciousness arising from applied statistics".

Real problems we face with AI are:

Ghost labor, erosion of the rights of artists, costs of automation, the climate impact of data-centers and the human impact of biased, opaque, incompetent and unfit algorithmic systems.

https://pluralistic.net/2023/11/27/10-types-of-people/

34 comments

"ai" as people think of it is marketing spin.
- Yeah, we had to rename AI to AGI, because marketing fuckers decided they're naming a (very smart) predictive model an AI. I've had a dumber version in my phone decades ago, this should've never been called AI.
  
  AI and AGI have never been the same thing.
  
  Whenever some dipshit responds to me with "you're talking about AGI, this is AI", my only reply is fuck right off.
"This article was sponsored by Roko's Basalisk/The paperclip maximiser"
- Capitalism is the paperclip maximizer, don't need AGI for that. AI fears are a tool it uses to achieve maximization.
  
  You clearly don't understand what the paperclip maximiser is. It doesn't just produce a lot of paperclips. It doesn't just kill most people. It's kills everyone and turns the entire planet into paperclips. And if we ignore AI safety by saying 'nOt My JoBs', we're just letting it happen.
Obviously, trivially, blatantly false, because the AI safety people have been at it since long before there was anything to market. Back then, the bullshit criticism was "AI will never be able to understand language or interpret pictures; what harm could it possibly ever do?"
- AI still doesn't "understand" language or pictures or anything else. It's little more than statistical analysis based on text and input from humans tagging photos. The fact that we can get some neat output is not indicitive that any understanding is going on behind the scenes.
- Even today there are a ton of people who simply seem uncapable of playing out the thought experiment about AGI being more competent than humans at virtually everything. They somehow seem to imagine our current generative AI models somehow being a proof that it can never actually deliver and become what the AI safety people have been worried about for decades.
- Yeah and now it's just people fearmongering "AI is the worst thing to happen to artists, musicians, graphic designers - and it steals everyone's work! And and it might become Skynet! And and.." people who don't understand technology always seem to do this. They did it with Crypto, did it with NFTs, and now legitimate technologies have to escape a reputational black hole all because scammers and grifters decided to use it maliciously.
  
  people who don’t understand technology always seem to do this. They did it with Crypto, did it with NFTs
  😬
  
  "Those morons warned us about these other things (that definitely did have all the problems that they warned about, nothing more annoying than a moron that's repeatedly proven right. What complete idiots.) Anyway they're definitely wrong about this thing."
  Do you hear yourself bruv?
  I have mixed feelings about "AI" , and you've practically convinced me to be against it. I doubt you could craft a less compelling argument if you tried.
"AI" has been diluted to mean nothing but marketing wank. 99.999% of "AI" is single variable linear regression estimation. Almost all of the rest is multi variable linear regression.
- Computers are a scam too. I found out they're nothing but ones and zeros!
itt people who under no pretext will read the linked article

34 comments