Matthew: My belief on this is we wanted AGI, we got AGI, we learned it is our brain that is still limited by sleep and distraction and whatever. And so I personally believe there are no limitations on the tech itself. There are massive limitations in the energy created to compute, the energy created for you and I to run the AI.
It is the typical genie analogy. Hey Matt, what do you want? I want to be tall. Alright, you are now ten feet. Whoa, whoa, AI, you overoptimized. I really wanted to be six one. Well, you should have told me you wanted to be six one. So it is a human problem now.
Lee Cage Jr.: More specific. And it is ask better questions and give better answers too. So it is even in prompting. I have looked at some prompt engineering courses and I have done this YouTube deep hole of like a hundred videos on having better prompts, asking better questions, being more specific.
But the thing is, what I like about the prompting is it has helped me figure out the gaps in my own personal communication. And the assumptions that I have placed on others when I have communicated to them.
So how often do you say, yeah, just add cheese? And then you get this chipotle salad that has a mound of cheese, and you are like, no, no, I wanted light cheese. Well, you did not say that. When you are prompting, it requires you to be as specific as possible to get the outcome that you are imagining.
I do think the limitations of the system are almost limitless. But there are some environmental, some ethical, and some moral concerns about AI as a whole.
