Honestly, the problem when talking about "AI" is how many different things that can mean.
General AI chats
Coding agents
Automated pentesting/vulnerability discovery
Image/video/music generation
Grammar checking
Automated support agents (phone or chat)
Autonomous weaponry
and so many more. Being Pro-AI could mean you like one or two application of the AI, but be against it in the others. I know very few people that like it for the use of media generation. However, there have been a lot of long time vulnerabilities in very popular open source projects that was only just discovered. That seems like a pretty undeniable use case demonstrating its usefulness.
Then of course there's governments that want to get their greedy blood thirsty hands on it to create autonomous weaponry. So now if you try to defend AI for a use case like defensively finding program vulnerabilities you somehow also have to defend AI weaponry?
For a generic AI model, it is very powerful and can either be used to grow yourself or abused so your brain doesn't have to work at all. You can use AI to do the hard work for you, or use it as a personal tutor to guide you into what to learn. People will of course mention hallucinations as why it can't be used to learn, but you don't have to take AI at its words. If you were to ask it to create a lesson plan on what you should study for a subject, in what order, and resources are available, you can do all of the actual learning using content the AI has no control over. So what you do with that is going to be up to the person, and opinions on it are going to vary wildly.
Some people argue any use case is not okay given the various concerns of energy and water usage, and where those models sourced their training data. Not to mention if you support AI you must be supporting the AI companies. I agree there are concerns for the environmental impact, and the training data discussion is a long one on its own. However, I do think you can support AI as a technology, and not be okay with the way the technology is being done in regards to environmental impact. And given AI can be done on a local machine, I don't think it has to be tied at all with the big tech at all.
"AI" is such a wide and immense topic. And what we talk about with AI today will not be relevant come next year with how quickly it is developing. We shall see if some form of Moore's law applied with the growth of AI as far as efficiency and quality of the AI goes.
Their concerns about drinking water are exaggerated, but technically your computer does use water if go far enough back. But once you do that, everything that uses electricity also uses water.
It's not nearly as much water as using ethanol gas.
If you point out any valid use case for AI, anti-AI people will come down your throat for sucking off Jensen Huang and promoting genocide.
There is also a difference in argument about what will become of AI. Anti-AI argue it is just a fad like crypto, and will die out. Others saying this is just the beginning and the technology will grow and revolutionize various industries.
So even if you don't like AI and AI datacenters, even acknowledging the fact AI is almost certainly the next massive revolution in technology is enough for people to start going down a rabbit hole.
AI hs already been demonstrated as a tool that largely benefits fascists and oligarchs.
Lmfao, what? The internet is also a tool that largely benefits fascists and oligarchs. Does that make every user of the internet a fascist, or just stupid?
Of course bad actors are going to take advantage of a tool that is very useful in an absurd amount of contexts...
It entirely depends on if the parents being well-off allowed their kid to get a degree they aren't really qualified for because they could use money as a crutch.
it only generates the most statistically likely text string to follow
Sure, but if that looks identical to what an ectual expert would say, is there really a difference?
LLMs have no expertise, it can only regurgitate text strings that are most likely, based on stealing the creative and scientific output of actual experts
I do agree there is a problem eith AI companies using copyrighted works for training, but couldn't you say the same thing about humans? If I read a book and tell someone else what I learned from that book, did I steal that information? If I pirated the book, yes. If I bought the book, then most people woulf say no.
as long as LLM providers decide they provide that
You say this as if we cant self host AI models. The companies don't matter given this fact
Anyways, you read like a giant AI-Booster
I can separate greedy tech corporations from the development of new technology. If you denied electricities expansion because the Edison company was greedy, you'd be fighting the existence of electricity in our infrastructure.
But it’s ok, I can just block you for your “Mein AI-Kampf” screed here.
That's probably for the best, given how that is an absurd ad hominem that is not related whatsoever to our discussion.
My main argument is that recognizing AI as a technology is inevitable and revolutionary, is very different from being okay with the AI mega corps, how they are devloping AI, and the environmental impacts of it. AI can be done right, that doesn't mean it is.
Blockchain was supposed to be very good for a handful of things. Smart contracts were never going to revolutionize the world, maybe a couple industries. I did not care about block chain technology, it was too specific.
I can have a personal tutor give me a customized lesson plan for little to no cost. It can look at large amounts of code in little to no time.
Energy comes from other sources than oil and methane. The fact the US is not prioritizing those alternate methods is a different problem.
AI will get more efficient, it will get better with less energy and storage. I am not saying the tech bros are not problematic. I am saying their problems do not negate the reality of AI as a technology, and the things it has unlocked.
Nuclear fusion created both atomic weapons and nuclear energy. Somebody advocating for nuclear energy is not also advocating for more nuclear warheads.
Cryptocurrency and NFTs are both blockchain. Blockchain by definition have very limited applications. It is hyped because anyone can use it to rug pull others pretending to be an investement. I'm not putting my life savings in AI. Not only that, but how can I use blockchain to help me with my schooling or cooking or thinking through a conplex problem? Metaverse is a product from a single company, not a technology.
AI as an idea would be more comparable to electricity, internet, smart phones, or other widespread technological advancement. They impact basically every industry, both residential and commercial applications. They are a technology and not specific to a single company (though there are some dominant players). This is more like an advancement in a game that has just been unlocked. There is no re-locking it.
It is inevitable because even if everyone stopped making more, people would still be using it forever. It is already too good to not use. And it will only get better. That could not be said for your other examples. How is that anything but inevitable?
I think fighting the tech companies makes sense. But fighting AI as a technology does not make sense.
Comparing AI to the guillotine is absurd. The guillotine and beheadings would literally only impact 1 single industry, and no other.
At one point adding electricity meant adding to the bottom line of Thomas Edison and Edison Illuminating Company. Why spend all the money and infrastructure when you are just going to get light when candles work fine? Not to mention how dangerous and ugly electric poles are. Who wants those hovering over their head??
People literally fought electricity because they were looking too close. If you take a step back, you know when something is a big deal based on what potential impacts it could have on a variety of industries. Same with the internet. People saw the potential, and corporations saw the money. There was a bubble, the bubble popped, and the real potential was left behind. AI bubble will pop at some point, and the genuine application of AI will be left behind and whether we want it to or not, will change the world. The technology exists, and pandoras box has been opened.
This is not a matter of faster beheadings, this is a matter of a conversational pocket expert with immediate response times at any time of day. The technology will only get better, it will only get more efficient. Nvidia will not hold their power forever. You can separate the technology from the corporations profiting and recognize what will happen.
How is it invalid? You are missing out on optimizations if you arent programming in assembly. We don't anymore because technology has changed, and tools have been developed to make writing code easier and faster.
As part of our continued collaboration with Anthropic
Anthropic is literally the one that refused to let them make autonomous weapons with their AI. There is a whole wikipedia page about it. They explicitly don't want their AI used for weapons. Of course, that wouldn't stop governments/militaries from doing so anyway. It would be different if Mozilla was working with OpenAI, but of the two Anthropic is currently the better one.
And yes, the AI is out of the box. Just like once nuclear warheads were created, there is no going back.
No? Things just will exist once they are discovered. Once nuclear warheads were discovered, there was no going back. Once the internet was established, it was going to be around. Same goes for a million other things. And now it applies to AI as well. Even if every big tech fascist stopped making AI, it is still going to be around, and it will be used maliciously. Our best bet is to use it defensibly before it can be used against us.
Occasional jokes in a topic filled with mostly serious answers is not problematic, especially ones that aren't at the expense of anyone else. I think the joke was appropriate. Especially on a platform that is lucky to get 30+ comments on a post, it being top comment is not going to overshadow the others. Most people can read all of the comments in a few minutes anyway.
Honestly, the problem when talking about "AI" is how many different things that can mean.
and so many more. Being Pro-AI could mean you like one or two application of the AI, but be against it in the others. I know very few people that like it for the use of media generation. However, there have been a lot of long time vulnerabilities in very popular open source projects that was only just discovered. That seems like a pretty undeniable use case demonstrating its usefulness.
Then of course there's governments that want to get their greedy blood thirsty hands on it to create autonomous weaponry. So now if you try to defend AI for a use case like defensively finding program vulnerabilities you somehow also have to defend AI weaponry?
For a generic AI model, it is very powerful and can either be used to grow yourself or abused so your brain doesn't have to work at all. You can use AI to do the hard work for you, or use it as a personal tutor to guide you into what to learn. People will of course mention hallucinations as why it can't be used to learn, but you don't have to take AI at its words. If you were to ask it to create a lesson plan on what you should study for a subject, in what order, and resources are available, you can do all of the actual learning using content the AI has no control over. So what you do with that is going to be up to the person, and opinions on it are going to vary wildly.
Some people argue any use case is not okay given the various concerns of energy and water usage, and where those models sourced their training data. Not to mention if you support AI you must be supporting the AI companies. I agree there are concerns for the environmental impact, and the training data discussion is a long one on its own. However, I do think you can support AI as a technology, and not be okay with the way the technology is being done in regards to environmental impact. And given AI can be done on a local machine, I don't think it has to be tied at all with the big tech at all.
"AI" is such a wide and immense topic. And what we talk about with AI today will not be relevant come next year with how quickly it is developing. We shall see if some form of Moore's law applied with the growth of AI as far as efficiency and quality of the AI goes.