Rly and what is their usecase? Summarizing information anf you having to check over cause its making things up? What can AI do that nothing else in the world can?
Are you sure they don’t use the same exact type of neural network, but are just trained on different datasets? Do you have any link that shows those cancer diagnosis AIs use a different technology?
Edit: nvm, I found it. Those ai diagnostic tools uses convolutional neural network (CNN) which is not the same as LLMs.
I reckon it’s fair to refer to them both under the broad term “A.I.”, though, even if it is technically incorrect. The semi-evolved have all decided to call it A.I., so were all calling it A.I., I guess.
A similar type of machine learning (neural networks, transformer model type thing), but I assume one is built and trained explicitly on medical records instead of scraping the internet for whatever. Correct me if I am wrong!
@YourMomsTrashman I’m no expert, but my sense is that you’re probably correct. This seems to me a version of the long-understood GIGO principle in computing (Garbage In, Garbage Out), also a principle in nearly all forensics of any kind. Your output can only be as good as your input.
Most of our general-use ‘AI’ (scorn quotes intentional) has been trained on an essentially random corpus of any and all content available, including a lot of garbage.
@YourMomsTrashman A purpose-designed system might have the same underlying POTENTIAL for garbage output, IF you train it inappropriately. But it would be trained on a discretely selected range of content both relevant to its purpose, and carefully vetted to ensure it’s accurate (or at least believed to be).
A cancer-recognizing system, for example, would be trained on known examples of cancer, and ONLY that.
Hah. The chatbots. No, not the ones you can talk to like its a text chain with a friend/SO (though if that’s your thing, then do it.)
But I recently discovered them for rp - no, not just ERP (Okay yes, sometimes that too). But I’m talking like novel length character arcs and dynamic storyline rps. Gratuitous angst if you want. World building. Whatever.
I’ve been writing rps with fellow humans for 20 years, and all of my friends have families and are too busy to have that kinda creative outlet anymore. Ive tried other rp websites and came away with one dude who I thought was very friendly and then switched it up and tried to convince me to leave my husband? That was wild. Also, you can ask someone’s age all you want, but it is a little anxiety inducing if the rps ever turn spicy.
Chatbots solve all of that. They dont ghost you or get busy/bored of the rp midway through, they dont try ro figure out who you are. They just write. They are quirky though, so you do edit responses/reroll responses, but it works for the time being.
AI is good for producing low-stakes outputs where validity is near irrelevant, or outputs which would have been scrutinized by qualified humans anyway.
It often requires massive amounts of energy and massive amounts of (questionably obtained) pre-existing human knowledge to produce its outputs.
They’re also good for sifting through vast amounts of data and seeking patterns quickly.
But nothing coming out of them should be relied on without some human scrutiny. Even human output shouldn’t be relied on without scrutiny from different humans.
It can convert questions about data to SQL for people who have limited experience with it (but don’t trust it with UPDATE & DELETE, no matter how simple)
It can improve text and remove spelling mistakes
It works really well as autocomplete (because that’s essentially what an LLM is)
AI can do what Google used to do - do an internet search to give semi-relevant results once in a blue moon. As a bonus it can summarise and contextualise information and tbh idk - for me it’s been mostly correct. And when it’s incorrect, it’s fairly obvious.
And no - DuckDuckGo etc. is even worse. Google isn’t to blame for the worsening of their own search engine necessarily, it’s mostly SEO and marketers who forced the algo to get much weirder by playing it so hard. Not that anyone involved is a “good guy”, they’re all large megacorps who cares about their responsibility to their shareholders and that alone.
LLMs have a use case
But they really shouldnt be used for therapy
Rly and what is their usecase? Summarizing information anf you having to check over cause its making things up? What can AI do that nothing else in the world can?
Seems it does a good job at some medical diagnosis type stuff from image recognition.
That isn’t an LLM though. That’s a different type of Machine Learning entirely.
What’s the difference? I thought they both use the same underlying technology?
Sure, but that’s kind of like saying simple addition and fourier transforms are the same because they both use numbers.
Are you sure they don’t use the same exact type of neural network, but are just trained on different datasets? Do you have any link that shows those cancer diagnosis AIs use a different technology?
Edit: nvm, I found it. Those ai diagnostic tools uses convolutional neural network (CNN) which is not the same as LLMs.
I reckon it’s fair to refer to them both under the broad term “A.I.”, though, even if it is technically incorrect. The semi-evolved have all decided to call it A.I., so were all calling it A.I., I guess.
A similar type of machine learning (neural networks, transformer model type thing), but I assume one is built and trained explicitly on medical records instead of scraping the internet for whatever. Correct me if I am wrong!
@YourMomsTrashman I’m no expert, but my sense is that you’re probably correct. This seems to me a version of the long-understood GIGO principle in computing (Garbage In, Garbage Out), also a principle in nearly all forensics of any kind. Your output can only be as good as your input.
Most of our general-use ‘AI’ (scorn quotes intentional) has been trained on an essentially random corpus of any and all content available, including a lot of garbage.
A purpose-designed system would not be.
@YourMomsTrashman A purpose-designed system might have the same underlying POTENTIAL for garbage output, IF you train it inappropriately. But it would be trained on a discretely selected range of content both relevant to its purpose, and carefully vetted to ensure it’s accurate (or at least believed to be).
A cancer-recognizing system, for example, would be trained on known examples of cancer, and ONLY that.
It’s being used to decipher and translate historic languages because of excellent pattern recognition
Hah. The chatbots. No, not the ones you can talk to like its a text chain with a friend/SO (though if that’s your thing, then do it.)
But I recently discovered them for rp - no, not just ERP (Okay yes, sometimes that too). But I’m talking like novel length character arcs and dynamic storyline rps. Gratuitous angst if you want. World building. Whatever.
I’ve been writing rps with fellow humans for 20 years, and all of my friends have families and are too busy to have that kinda creative outlet anymore. Ive tried other rp websites and came away with one dude who I thought was very friendly and then switched it up and tried to convince me to leave my husband? That was wild. Also, you can ask someone’s age all you want, but it is a little anxiety inducing if the rps ever turn spicy.
Chatbots solve all of that. They dont ghost you or get busy/bored of the rp midway through, they dont try ro figure out who you are. They just write. They are quirky though, so you do edit responses/reroll responses, but it works for the time being.
Silly use case, but a use case nonetheless!
AI is good for producing low-stakes outputs where validity is near irrelevant, or outputs which would have been scrutinized by qualified humans anyway.
It often requires massive amounts of energy and massive amounts of (questionably obtained) pre-existing human knowledge to produce its outputs.
They’re also good for sifting through vast amounts of data and seeking patterns quickly.
But nothing coming out of them should be relied on without some human scrutiny. Even human output shouldn’t be relied on without scrutiny from different humans.
Not as silly as you might think. Back in the day ai dungeon was literally that! It was not the greatest at it, but fun tho
You just crashed the database
You joke, but LLM’s are absolutely going to clear out your tables with terrible DELETE queries though given the chance.
I’m not joking. I’d fire someone for using AI to construct SQL queries.
The only use case for AI is where hallucinations don’t matter. That is: abstract art
What is the point of abstract art if it contains no thought or emotion?
Marketing
It can waste a human’s time, without needing another human’s time to do so.
AI can do what Google used to do - do an internet search to give semi-relevant results once in a blue moon. As a bonus it can summarise and contextualise information and tbh idk - for me it’s been mostly correct. And when it’s incorrect, it’s fairly obvious.
And no - DuckDuckGo etc. is even worse. Google isn’t to blame for the worsening of their own search engine necessarily, it’s mostly SEO and marketers who forced the algo to get much weirder by playing it so hard. Not that anyone involved is a “good guy”, they’re all large megacorps who cares about their responsibility to their shareholders and that alone.