Raw LLM Responses

I have to say not my favorite Ted talk. I disagree with a few things, for example, carbon footprints And polluting the environment by using a computer seriously.? People will be using their computers regardless of AI existing or not. At least some people are using AI to help them create or discover things or solve problems ... such as saving the environment etc. But I digress. If AI some day replaces lawyers or Certain doctors like radiologists, Etc, well, these people won't be driving cars to work all the time and polluting. So in a sense AI is doing the reverse is it not ? And regarding AI being racist or sexist and all that. I think the data that is input into these large language models needs to be as complete as possible. That way, if you ask a model to generate an image of a scientist, Or of a criminal...it will show people of all races, sexes, ages etc. You also have to keep in mind when asking AI to generate such things. The more precise you are, the more precise the outcome will be, such as asking to generate an image of a female scientist wearing glasses with long red hair and a white lab coat VS just asking it to Generate a scientist or a CEO.... if asked to Generate a criminal or a terrorist, you will probably get the same thing over and over again, without a complete data set for AI to pull from...What if you say generate, A cannibalistic pedophile? You will probably always get an image of a white male... because you can't argue with statistics and true data? AI is kind of still in its Infancy, so it still needs time to continue to learn. And it's up to us to be precise when asking it to perform something or generate something. And it's up to us humans to input as much accurate data as possible. To get accurate results as an output. Someday it might have so much information that we could ask it to generate a cure for a horrible disease... To me that's worth all the carbon in the world is it not?