I feel the same way about AI that I did about crypto when it started becoming a thing:
• All of the moral practices that make it possible are fucked beyond repair.
• It’s spawning some of the worst companies, and enriching some of the worst billionaires in the world.
• While there are some genuinely useful, practical applications for it, there are so few of them, and they have had so little positive impact so far, that it doesn’t even seem like it’s going to be worth it anytime soon.
@nileane Ai is not new, and everything you've written shows your ignorance of the subject.
Ai is not just chat bots, it's climate analysis and modeling, vaccine and antiviral research, global and local infrastructure automation and realtime data analysis, it's in astrophysics and solar flare prediction science, etc... it's everywhere that people like you never seem to care about.
True, but all the AI models in cancer, climate, ... research are fundamentally different from the ChatGPT type AI that is hyped at the moment. And the tech bros are less interested in those useful forms it seems.
I'd argue that it is part of the unethical practice, though, to conflate the two and attempt to disarm any criticism of the ChatGPT type AI pointing at AI use in cancer and medical research.
And it it indeed a shame if those uses get a bad reputation as a fallout from all the everyday AI hype.
@nileane the timing is also not helping. Back in 2016 a lot of us realized how powerful and how fucked up big tech can be and the impact it can have in our world, misinformation being a big thing!
Did we try to solve this? No, we invested in AI where it's like throwing gasoline into a fire where we have no way of knowing if anything is true or not. What could go wrong?
@nileane the one thing that I think will be a positive outcome from all of this, is that the general public will finally twig how valuable their data is.
One can wish. But thru my lifetime, I've seen how we have gladly offshored manufacturing for the product of cheap labor, cheap goods. And until the recent (DEMOCRAT) recognition that we'd cut too deeply into essential industries, and re-establishment of some industries here, it was a downward spiral of trade deficit and weakening security.
AI is in an analogous position. As long as it produces 'cheap crap' that people like because of the immediacy and (false) economy, no real guardrailing is going to take place.
Externalizing cost and risk is the hallmark of capitalism. Taking stuff at one cost (trees, oil, data, land, other peoples money, etc) then selling it for profit, is basically Econ101. The evil part is how egregiously the costs can be avoided, externalized, or as in the case of fossil fuels... delayed for a few generations...
AI as a concept is being oversold. I think the reality of it is it is a natural progression in machine learning. We're in a position right now where its most visible applications to the public are online content generation, 6 fingers and all.
My son's in Silicon Valley in the machine learning biz, and we talked about this 12 hours ago, coincidentally. Even the projects he is working on, although they use AI/ML, are not directly interfaced with other parallel projects in the company. The machines are not yet scraping other machines. That's when the real fun begins.
But right now, AI's primary usefulness in the market seems to be pumping AI stock prices. From the cheap seats, where I sit, at least.
@skydog@oldguycrusty@nileane@sigmund Agreed. I've worked a bit with ML, and followed it for years. It's very useful tech, with a lot of known limitations and issues.
The current boom of generative "AI" (which isn't AI at all) is largely a money grab and stock bubble at the expense of privacy and energy, and a boom to fraud and misinformation at a time in our history where it can do a lot of social damage. As if the tech bros ever cared about the consequences.
Add comment