Personally what I hate is not the tech developments being labelled “AI”. It’s the industry behind it, and how much it filths itself with deception.
This sort of neural network is good for small and menial tasks, where accuracy is not too important but volume is. For that you don’t need large models, you need smaller ones, that take a fraction of the data and energy to process (“train”). Then you’d advertise them for what they are - a bunch of useful tools.
But we’re talking about an industry led by con artists, billionaires, liars and vulture capital. Their eyes get bloody in rage, if they don’t see smoke and mirrors; they don’t care about truth, but appearances. It needs to look “grandiose”, it needs “hype”, it needs “marketability”. It needs all that “AGI SOON!”.
So the models get bigger, bigger, and bigger. But not necessarily better; more sycophant, more assumptive, more energy-demanding.
Then you plug everything wrote in the article as a consequence.
Personally what I hate is not the tech developments being labelled “AI”. It’s the industry behind it, and how much it filths itself with deception.
This sort of neural network is good for small and menial tasks, where accuracy is not too important but volume is. For that you don’t need large models, you need smaller ones, that take a fraction of the data and energy to process (“train”). Then you’d advertise them for what they are - a bunch of useful tools.
But we’re talking about an industry led by con artists, billionaires, liars and vulture capital. Their eyes get bloody in rage, if they don’t see smoke and mirrors; they don’t care about truth, but appearances. It needs to look “grandiose”, it needs “hype”, it needs “marketability”. It needs all that “AGI SOON!”.
So the models get bigger, bigger, and bigger. But not necessarily better; more sycophant, more assumptive, more energy-demanding.
Then you plug everything wrote in the article as a consequence.