Seems like the perfect technology to create a browser extension that re-writes article headlines to be non-clickbait and actually reflect the content of an article. Bonus points if it could redo youtube titles and thumbnails.
There are plenty, as others have said, it has been able to help a lot in science. It has been used for finding new drugs, search for composite materials.
Sadly i don’t think the pros will outweigh the cons, the society as is right now won’t let it happen.
I truly think this tool, even without any fancy AGI, if used in a responsible manner could help us solve some of the most pressing problems the world has.
But it won’t happen, look at the irresponsible manner the capitalist society has rolled out this tech, which is still in its infancy. It has wasted tons and tons of energy on models that are not yet fully mature. Brute force training of shit models just to get the best monthly benchmark over competitors.
It’s so sad to see all the promised goals in terms of saving energy and water that went to garbage the moment the AI hype exploded.
And what we got for that? A bunch of models causing problems because they are not ready. Executives cutting workforce, in part to justify the use the new shiny tooling, only to get shittier services. And worse of all, shortage of water and energy to pump more compute.
Even if we get to a future where AI is cheap, powerful and capable, it will only be used to widen the gap between the wealthy and the common people.
****Stephen Hawking said it way more eloquently already in 2016****
The potential benefits of creating intelligence are huge. […]. Perhaps with the tools of this new technological revolution, we will be able to undo some of the damage done to the natural world by the last one — industrialization. And surely we will aim to finally eradicate disease and poverty. […] In short, success in creating AI, could be the biggest event in the history of our civilization.
But it could also be the last, unless we learn how to avoid the risks. Alongside the benefits, AI will also bring dangers, like powerful autonomous weapons, or new ways for the few to oppress the many. It will bring great disruption to our economy.
Tons. I think if people read my opinions on AI they likely see me as a luddite.
My concerns are not about whether it’s useful. It’s that if the 1% use it to replace most actual workers, the lack of input will make future models actually worse than current ones, and at the very least would stifle innovation.
I’m very concerned about models built on the IP (voluntarily given or not) of people, being used to replace those same people.
I’m very concerned with where we go as a society if we do go down the route of losing so many jobs.
I’m concerned about the race to get the best model, using so much energy and natural resources.
But do I think AI is and can be a very powerful tool, to enhance productivity? Without a doubt it can.
I use the RADE V1 codec in FreeDV. It is an AI vocoder that packs high quality audio into 1500 Hz of bandwidth. It sounds much better than traditional codecs like AMBE or Codec 2.
Another one I use is AI motion detection for my security cameras. It recognises a vehicle or a person and significantly reduces the number of false alerts.
My Reolink cameras have pretty good AI detection, I think that is a decent use of AI if it runs locally. I like seeing the “Animal” detections from my doorbell camera, it’s usually a cat or a rabbit investigating the front steps.
Mine runs locally on the DVR. I wouldn’t trust any cloud service with my video.
My old setup only had normal motion detection. I couldn’t enable alerts with that because it would go off all night from the bugs swarming the IR LEDs on the cameras.
Of course there are many legitimate and/or constructive uses for AI. Like any new technology, it depends on what you do with it. You can use it do help or harm, that is up to us.
For example, in many countries getting a radiologist to look at medical scans takes a lot of time and is a huge bottleneck. AI can help accelerate this.
Are þey worþ þe cost? If we count þe real cost - not þe subsidized, but þe hidden, þe environmental, þe social, and þe intellectual costs - do þey add equivalent value?
Are þey eþically justifiable given þat þey are plagiarism machines?
Yes, they tend to be pretty good at pattern recognition, which has enormously useful applications in manufacturing, data sorting, translation, and more. There are absolutely use cases for AI/LLMs. The problem is that, under capitalism, LLMs are being overhyped as a new market for finance capital to saturate (creating a bubble), and that these huge, inefficient data centers are being pushed as ultimate investment vehicles at the immense destruction of the environment. In a socialist economy, with strong guardrails on the development of LLMs, these can be made to help society in general. See how it’s treated in China vs. the US Empire.
AI/LLMs are not a panacea, despite what investors will try to tell you. At the same time, they are not utterly worthless. They have definite use cases that will be found gradually.
It’s a tool. As engineer, I will try to use each and every tool that helps me, so it’s reasonable for me to learn using this tool effectively as well: what works, what doesn’t, what’s expensive, how can I keep it under control etc.
Of course thats only the purely technical/utilitarian aspect, and I’m not going into the moral/social aspects here.
LLM as a glorified search engine seems kind of ok. It’s the generative side that is awful. And the search engine aspect might do harm than good, since it’s also used for mass surveillance.
Machine translation is now working fairly well, maybe just in time. Within a decade or so, most technical and scientific publications will probably be in Chinese. So the translation tools will help us backwards English speakers read them.
Personally, I decided using a local LLM was acceptable for generating translation strings in my open source application. It had already been manually translated into a bunch of different languages by various contributors over the years, but I just did a major UI rework and so most of the existing translations now had a ton of missing entries. Qt6’s translation tool had an AI Translation function so I decided to try it out after setting up ollama and the recommended qwen3 model. It did a pretty decent job as far as I can tell, I am only fluent in English but I did some cross checking by translating back to English via Google Translate and it seemed to do a decent job. It at least got all the strings translated to something usable, once it’s merged then I expect native speakers will contribute cleanups and rewordings if needed.
I don’t really consider translation purely generative as it is more of a conversion task. I think AI is pretty useful for this. Same for things like automatic captioning and even mundane text to speech (if not being used to replace voice acting).
I think as long as there is competent human verification of what it’s done, it’s legitimate to use AI for nearly anything. Especially the creation of art.
But I’ve also read that AI has been helpful in finding security problems in software that humans likely wouldn’t have found. I don’t see a reason to hate AI in general.
No, and I’m tired of hearing legitimate and important criticisms of LLMs and the whole AI industry start with “There are things AI is good for but…”
There is so much wrong with this technology, how it was created, how it’s being pushed, how it’s being funded, how it affects us, and all of the necessary discussion is being pre-emptively watered down by this prevarication of critics, just to pander to the AI-brained robot-fuckers that can’t bear to hear that their favourite toy that allows them to role-play as being competent is just a piece of abusive shit designed to make us stupider and trap us in an endless dependency on the new oligarchs’ expensive infrastructure while giving them a convenient scapegoat for all of their malicious acts, propaganda and history rewriting.
There is nothing that LLMs are good for, and every time you make excuses for it, you’re tightening the noose around your own neck, and all of the people that come after us.
You are not logged in. However you can subscribe from another Fediverse account, for example Lemmy or Mastodon. To do this, paste the following into the search field of your instance: [email protected]
This is the official technology community of Lemmy.ml for all news related to creation and use of technology, and to facilitate civil, meaningful discussion around it.
Ask in DM before posting product reviews or ads. All such posts otherwise are subject to removal.
Rules:
1: All Lemmy rules apply
2: Do not post low effort posts
3: NEVER post naziped*gore stuff
4: Always post article URLs or their archived version URLs as sources, NOT screenshots. Help the blind users.
5: personal rants of Big Tech CEOs like Elon Musk are unwelcome (does not include posts about their companies affecting wide range of people)
6: no advertisement posts unless verified as legitimate and non-exploitative/non-consumerist
7: crypto related posts, unless essential, are disallowed
Seems like the perfect technology to create a browser extension that re-writes article headlines to be non-clickbait and actually reflect the content of an article. Bonus points if it could redo youtube titles and thumbnails.
There are plenty, as others have said, it has been able to help a lot in science. It has been used for finding new drugs, search for composite materials.
Sadly i don’t think the pros will outweigh the cons, the society as is right now won’t let it happen. I truly think this tool, even without any fancy AGI, if used in a responsible manner could help us solve some of the most pressing problems the world has.
But it won’t happen, look at the irresponsible manner the capitalist society has rolled out this tech, which is still in its infancy. It has wasted tons and tons of energy on models that are not yet fully mature. Brute force training of shit models just to get the best monthly benchmark over competitors. It’s so sad to see all the promised goals in terms of saving energy and water that went to garbage the moment the AI hype exploded.
And what we got for that? A bunch of models causing problems because they are not ready. Executives cutting workforce, in part to justify the use the new shiny tooling, only to get shittier services. And worse of all, shortage of water and energy to pump more compute.
Even if we get to a future where AI is cheap, powerful and capable, it will only be used to widen the gap between the wealthy and the common people.
****Stephen Hawking said it way more eloquently already in 2016****
Link Article
Sorry for the doomer rant, its a conflicted topic for me, feel free to skip.
Tons. I think if people read my opinions on AI they likely see me as a luddite.
My concerns are not about whether it’s useful. It’s that if the 1% use it to replace most actual workers, the lack of input will make future models actually worse than current ones, and at the very least would stifle innovation.
I’m very concerned about models built on the IP (voluntarily given or not) of people, being used to replace those same people.
I’m very concerned with where we go as a society if we do go down the route of losing so many jobs.
I’m concerned about the race to get the best model, using so much energy and natural resources.
But do I think AI is and can be a very powerful tool, to enhance productivity? Without a doubt it can.
Sure, there are good uses.
I use the RADE V1 codec in FreeDV. It is an AI vocoder that packs high quality audio into 1500 Hz of bandwidth. It sounds much better than traditional codecs like AMBE or Codec 2.
Another one I use is AI motion detection for my security cameras. It recognises a vehicle or a person and significantly reduces the number of false alerts.
My Reolink cameras have pretty good AI detection, I think that is a decent use of AI if it runs locally. I like seeing the “Animal” detections from my doorbell camera, it’s usually a cat or a rabbit investigating the front steps.
Mine runs locally on the DVR. I wouldn’t trust any cloud service with my video.
My old setup only had normal motion detection. I couldn’t enable alerts with that because it would go off all night from the bugs swarming the IR LEDs on the cameras.
Of course there are many legitimate and/or constructive uses for AI. Like any new technology, it depends on what you do with it. You can use it do help or harm, that is up to us.
For example, in many countries getting a radiologist to look at medical scans takes a lot of time and is a huge bottleneck. AI can help accelerate this.
Yes. Right now I think the real questions are:
Yes, they tend to be pretty good at pattern recognition, which has enormously useful applications in manufacturing, data sorting, translation, and more. There are absolutely use cases for AI/LLMs. The problem is that, under capitalism, LLMs are being overhyped as a new market for finance capital to saturate (creating a bubble), and that these huge, inefficient data centers are being pushed as ultimate investment vehicles at the immense destruction of the environment. In a socialist economy, with strong guardrails on the development of LLMs, these can be made to help society in general. See how it’s treated in China vs. the US Empire.
AI/LLMs are not a panacea, despite what investors will try to tell you. At the same time, they are not utterly worthless. They have definite use cases that will be found gradually.
It’s a tool. As engineer, I will try to use each and every tool that helps me, so it’s reasonable for me to learn using this tool effectively as well: what works, what doesn’t, what’s expensive, how can I keep it under control etc.
Of course thats only the purely technical/utilitarian aspect, and I’m not going into the moral/social aspects here.
Machine Learning for pattern recognition in very large datasets seems like good usage.
LLMs are just auto-complete with sycophantic tendencies.
If and when AGI comes, it’ll evolve from robotics, and won’t be the egregious energy suck that the current brute force approach.
LLM as a glorified search engine seems kind of ok. It’s the generative side that is awful. And the search engine aspect might do harm than good, since it’s also used for mass surveillance.
Machine translation is now working fairly well, maybe just in time. Within a decade or so, most technical and scientific publications will probably be in Chinese. So the translation tools will help us backwards English speakers read them.
Personally, I decided using a local LLM was acceptable for generating translation strings in my open source application. It had already been manually translated into a bunch of different languages by various contributors over the years, but I just did a major UI rework and so most of the existing translations now had a ton of missing entries. Qt6’s translation tool had an AI Translation function so I decided to try it out after setting up ollama and the recommended qwen3 model. It did a pretty decent job as far as I can tell, I am only fluent in English but I did some cross checking by translating back to English via Google Translate and it seemed to do a decent job. It at least got all the strings translated to something usable, once it’s merged then I expect native speakers will contribute cleanups and rewordings if needed.
I don’t really consider translation purely generative as it is more of a conversion task. I think AI is pretty useful for this. Same for things like automatic captioning and even mundane text to speech (if not being used to replace voice acting).
There’s lots of great answers in this thread, but I’m still ready to start shooting at robots, anyway.
Plenty.
I think as long as there is competent human verification of what it’s done, it’s legitimate to use AI for nearly anything. Especially the creation of art.
But I’ve also read that AI has been helpful in finding security problems in software that humans likely wouldn’t have found. I don’t see a reason to hate AI in general.
No, and I’m tired of hearing legitimate and important criticisms of LLMs and the whole AI industry start with “There are things AI is good for but…”
There is so much wrong with this technology, how it was created, how it’s being pushed, how it’s being funded, how it affects us, and all of the necessary discussion is being pre-emptively watered down by this prevarication of critics, just to pander to the AI-brained robot-fuckers that can’t bear to hear that their favourite toy that allows them to role-play as being competent is just a piece of abusive shit designed to make us stupider and trap us in an endless dependency on the new oligarchs’ expensive infrastructure while giving them a convenient scapegoat for all of their malicious acts, propaganda and history rewriting.
There is nothing that LLMs are good for, and every time you make excuses for it, you’re tightening the noose around your own neck, and all of the people that come after us.
yup, it helps reducing wrist pain from typing