The Peculiar Persistence of the AI Denialists

https://www.persuasion.community/p/the-peculiar-persistence-of-the-ai

by sien • 12 months ago

6 5 12 months ago

d00mB0t• 12 months ago

AI Apologetics

willprice89• 12 months ago

It may be that the "denialists" persist and are so vocal because of the excessive hype, most of it marketing driven, around AI. Personally I think it's good to hear some counterpoints, even if both the denialists and the hypsters are way off the mark.

lordkrandel• 12 months ago

I'm sorry, you make a lot of points correct. It's true that AI will have impact. Translations, chess, medical analysis. Yet, for many many tasks, it is plainly unreliable. We always think of machine being autonomous 24/7 and perfect in their job. This is a new kind of machine. It may be good at translation, but I wouldn't send a translated message without checking it first. This is what hinders its supposed capabilities. To software developers, this is immensely evident. I can chat with it for hours, no way it's going to help me with my depression more than 30 minutes with my therapist. It can analyze MRI scans, but it doesn't know me or have an holistic approach to health. It may translate and hallucinate. It often throws nonsense in the slop. I would not read an AI book that is not redacted because the average of its training materials... is bad. And it's not a technology's fault, it's just a limitation. It is not fixable: it's the way the drill drills. This makes it fundamentally good for very limited and controlled aspects in our job. It cannot code, it cannot review, but yeah, it can probably help you learn a programming language better than a tutorial, given that you recognize its mistakes when they happen. Also, it cannot build upon itself, it doesnt really understand instructions. In the industry, it's about as helpful as a helpful dog. Maybe it will bring you the slippers and take out the garbage, but it's cute and eats a lot.

ggm• 12 months ago

History is the best judge. I was a kid during the lighthill report years, early 1970s, nothing makes me think it was wrong, and I expect future historians of science will look at this blip and say what they feel in retrospect about it.

50 years.

drweevil• 12 months ago

“AI won’t have that much impact, anyway.”

That's a bit of a strawman. One doesn't have to be pro-AI to see that it will have a large impact. Whether that impact is for the good or not is the issue. Just for starters, in a time that the impact of anthropogenic climate change is becoming ever more obvious the promise of the riches AI will bring to its investors has nonetheless resulted in setting aside any pretense of dealing with that existential threat. Instead we are now heedlessly rushing to build and use more and more power capability. To what end? To duplicate what people can do at least as well, if not better? And the damage that this will do to society, what of it? The investors don't care, and in this increasingly undemocratic environment the people affected will have little or no say. To me the AI craze epitomizes the general problem we are facing now: an out-of-control financial class that prioritizes their economic needs above all others, with the help of an entirely captured government. As if the economy were not something that must support us all.

I'm not looking forward to the fallout.

The Peculiar Persistence of the AI Denialists

Related Stories