Except, the other day I wanted to convert some units and the AI results was having a fucking stroke for some reason. The numbers did not make sense at all. Never seen it do that before, but alas, I did not take a screenshot.
Those LLMs can’t handle numbers, they have zero concept of what a number is. They can pull some definitions, they can sorta get very basic arithmetic to work in a limited domain based on syntax rules, but it will mess up most calculations. ChatGPT tries to work around it by recognizing the prompt is related to math, passing it to a more normal Wolfram-Alpha style algorithm, and then using the language model to format the reply into something more appealing, but even this approach often fails because if the AI gets confused for any reason it will feed moronic data to the maths algorithm.
LLMs don’t verify their output is true. Math is something where verifying its truth is easy. Ask an LLM how many Rs in strawberry and it’s plain to see if the answer is correct or not. Ask an LLM for a summary of Columbian history and it’s not as apparent. Ask an LLM for a poem about a tomato and there really isn’t a wrong answer.
The “sauce vs dressing” one worked for me when I first heard about it, but in the following days it refused to give an AI answer and now has a “reasonable” AI answer
I can literally never reproduce these. I’ve tried several times now.
Yeah, I never get these strange AI results.
Except, the other day I wanted to convert some units and the AI results was having a fucking stroke for some reason. The numbers did not make sense at all. Never seen it do that before, but alas, I did not take a screenshot.
Those LLMs can’t handle numbers, they have zero concept of what a number is. They can pull some definitions, they can sorta get very basic arithmetic to work in a limited domain based on syntax rules, but it will mess up most calculations. ChatGPT tries to work around it by recognizing the prompt is related to math, passing it to a more normal Wolfram-Alpha style algorithm, and then using the language model to format the reply into something more appealing, but even this approach often fails because if the AI gets confused for any reason it will feed moronic data to the maths algorithm.
What do humans do? Does the human brain have different sections for language processing and arithmetic?
LLMs don’t verify their output is true. Math is something where verifying its truth is easy. Ask an LLM how many Rs in strawberry and it’s plain to see if the answer is correct or not. Ask an LLM for a summary of Columbian history and it’s not as apparent. Ask an LLM for a poem about a tomato and there really isn’t a wrong answer.
Meanwhile, GNU Units can do that, reliably and consistently, on a freaking 486. 😂
That just might be fake
At one point they were packing a shitload of usb ports onto the IO panel, 5 stacks of 4 ports wouldnt surprise me
At least one dumb one was reproducible, I’d look for it but it was probably a few hundred comments ago
The “sauce vs dressing” one worked for me when I first heard about it, but in the following days it refused to give an AI answer and now has a “reasonable” AI answer
The original, if you haven’t seen it:
Love it :D