Yes, your initial question was indeed poorly worded.
I feel like you’re arguing semantics. Yes, technology exists to serve us. No, that doesn’t mean every tool needs to have some sort of use case checking, a tool will do as it’s used, provided it’s not a shit tool that explodes or whatever.
A machete will cut jungle or throats, it doesn’t know the difference or care, that’s on the user.
A plagiarism machine capable of no more than simply guessing at the desired output sold by charlatans as an artificial intelligence to an increasingly ignorant and desperate population is a much more complicated example.
It’s arguably working as designed, but ultimately the problem here is that of the aforementioned fraud. What they have is being sold as something it’s categorically not in the headline, with the reality buried in obfuscated small print.
The whole LLM bubble is balanced on the assertion that by scaling up the plagiarism machine, it will become greater than the sum of it’s parts. Now, enough advances in learning models have been made that AI research is only going to accelerate but the bubble around the current “product” isn’t economically sustainable. No matter what happens, there are going to be some massive losses when reality finally bites.
I wasn’t arguing semantics, i was trying to figure out what semantics you were using and why so i could better understand the perspective.
I agree with all of those LLM arguments, with the exception that the last one is missing a bit on the end.
The LLM bubble is working as designed, to enrich the people running the scam, lots of people are going to lose a lot of money but i’d wager on the architects/early participants of the scam coming out with a large net positive.
Yes, your initial question was indeed poorly worded.
I feel like you’re arguing semantics. Yes, technology exists to serve us. No, that doesn’t mean every tool needs to have some sort of use case checking, a tool will do as it’s used, provided it’s not a shit tool that explodes or whatever.
A machete will cut jungle or throats, it doesn’t know the difference or care, that’s on the user.
A plagiarism machine capable of no more than simply guessing at the desired output sold by charlatans as an artificial intelligence to an increasingly ignorant and desperate population is a much more complicated example.
It’s arguably working as designed, but ultimately the problem here is that of the aforementioned fraud. What they have is being sold as something it’s categorically not in the headline, with the reality buried in obfuscated small print.
The whole LLM bubble is balanced on the assertion that by scaling up the plagiarism machine, it will become greater than the sum of it’s parts. Now, enough advances in learning models have been made that AI research is only going to accelerate but the bubble around the current “product” isn’t economically sustainable. No matter what happens, there are going to be some massive losses when reality finally bites.
I wasn’t arguing semantics, i was trying to figure out what semantics you were using and why so i could better understand the perspective.
I agree with all of those LLM arguments, with the exception that the last one is missing a bit on the end.
The LLM bubble is working as designed, to enrich the people running the scam, lots of people are going to lose a lot of money but i’d wager on the architects/early participants of the scam coming out with a large net positive.