A reason I’m not afraid of AI. The question was does 21/7 + 1=4.
Both Alfred and Bard screw this up.
What’s interesting is they get DIFFERENT wrong answers.
Agreed. I'm finding these language modeler applications take more effort to prompt, than it does to do the work oneself.
They're over-hyped and over-glorified search engines.
Yes. They are overhyped because of one thing: money.
It will make them a shit ton of cash if people ignore the fact that they are AT BEST fallacious and AT WORST wrong.
No one buys a calculator that gives them the right answer 30% of the time.
The hype is for a product: except in specialized cases, AI is not a product. These deep learning models are fundamentally and functionally flawed.
The fact everyone is glossing over it drives me nuts.