What do you think Lemmy is most biased about? Which opinions do you think differ most from the general internet?
(Excluding US politics, due to community rules)
What do you think Lemmy is most biased about? Which opinions do you think differ most from the general internet?
(Excluding US politics, due to community rules)
Are people anti non-generative AI? Or is broader AI just getting dragged in to the justified anti genAI sentiment?
I think it’s both. Some people dislike all AI because of generative AI like LLM’s, but many people seem to care about making the distinction between generative AI and traditional ML.
I suspect a lot of the former group is don’t that out of ignorance or forgetfulness - I do it all the time, because I often assume people are talking about GenAI. Which is probably a reasonable assumption about 90% of the time these days, but it is better to be clear about it.
Also, a friend who has a background in AI draws a distinction between ML and non-generative AI: ML is basically tools for overpowered statistical analysis and pattern finding, AI is attempts to partially recreate aspects of intelligence, and can include evolutionary algorithms and stuff. Still not sure I see the distinction (and there is overlap), but they’re way more informed than me…
I formed this question to myself and was about to post it, but then I remembered Lemmy also hates self-driving cars which are likely Convolutional Neural Networks (CNN) or Recurrent Nerual Networks (RNN) which are not part of Generative AI at all.
I think the hatred there is completely disconnected from the fact that it’s AI.
Both of them have in common that the technology is being forced upon us at the cost of lives, livelihoods, and the environment upon which we all rely to survive.
I’m not sure I follow your logic. Those reasons you give are still hatred of AI because of those results (job loss, etc). How is that not hatred of AI?
I agree with Hawke, I think people are against the use of technology in such a way that it exploits workers and customers, not fundamentally against the technology itself.
Basically like the Luddites - they smashed weaving looms, not because the technology was fundamentally bad, but because it was being used by capitalists to worsen working conditions and destroy livelihoods.
I agree with that statement too. Where Hawke and I are disagreeing is I believe Self Driving cars can be used to exploit workers and customers. We already have Waymo robot taxi cabs that are displacing human drivers.
That’s not how I ready their statement… I think Hawke is saying that AI itself is not inherently bad, just that it’s being used for bad things. The bad things they identified are different from yours, but I think you’re basically saying the same thing over-all?
The self-driving cars are not hatred of AI. Nobody* cares that they use machine-learning to enable the cars to drive themselves.
It’s not hatred of AI there.
I still disagree, but let me create another hypothetical example that may highlight where we might disagreement further:
What if Deep Learning (not Gen AI) was used in missile guidance systems specifically to aim toward “people shaped targets”? Would the hate be for AI or just for missiles? If missiles is your answer, where is the distinction in your mind between that and the self-driving cars example?
GenAI being used in missile guidance makes zero sense - the technology is not applicable there, because you need precision and reliability. Normal AI, sure.
I specifically said “not Gen AI”.
Ah, sorry, I misread.
There is no relevant distinction.
I hate missiles for the damage they do regardless of the technology used to point them at the target.
I hate cars for the damage they do regardless of the technology used to point them at the target.
There’s a difference of intent (missiles get aimed at people while self driving cars hopefully get aimed away from people).
There’s a difference of failure modes (when a self-driving car fails it will often hit a person while a failing missile will miss a person).
But theres no reason to hate machine-learning for that, any more there is a reason to hate gyroscopes or lidar or other tools which are also used in guidance systems.
If someone had decided that a simple accelerometer were “good enough” to unleash self-driving cars on the general public without consideration for the damage caused, people would be upset and rightly so, but not because of the specific technology itself.
Edit: changed AI to “machine-learning”