But what problems are they solving for the average person? If they are so complicated, maybe they shouldn’t be mass marketed. Especially since they are a waste of energy for things basic web searches can do.
also, the average person finds the ti-89 useless not because of complexity but because they have no use for it. conflating the two ideas of uselessness (can't comprehend/don't fuckin' need) muddies the waters.
And, as I pointed out elsewhere, there is absolutely no one pushing for mass adoption of graphing calculators. I don't have one, I don't need one, I'm not being marketed to.
Not pro MaLIG* or LLM in general, but there are some niche applications for them if you do the opposite of what OpenAI and other hypergeneralist firms do with them and strictly control what data goes in.
That’s the thing. I’m not arguing that AI as a whole is useless. I’m saying that there are specific and more technical ways that it can be used - and it shouldn’t be marketed widely or expected to have huge use, especially at the expense of energy resources.