The biggest difference between LLMs and us, in terms of our thinking, is that LLMs are purely symbolic. They’re not capable of interacting with the world in any other way.
I’m pretty sure he didn’t study any sciences. He may have been in a classroom where studying of science was happening, but it wasn’t him doing it.
I also would like to point out that this is a very old saying. It has some wisdom to it, but it’s not meant to be taken literally.
My uncle was schizoaffective with severe psychosis and I think a pretty strong candidate for autism as well (it’s hard to tell because some antipsychotic meds can change your affect.)
He was a highly intelligent man - probably where my son got his raw math ability. He rambled a lot about unified fields and quantum physics and shit and I could never tell if he was talking about real science or just seeing patterns where none existed. I can only imagine what he’d be saying if he lived in the time of AI.
Actually, let me expand on that… I think that, in science classes, his main takeaway was “Man, this stuff is hard.”. Because it is. It’s so hard, in fact, that the vast majority of science can’t be comprehended, at all, by anyone. But the problem is, he took that as a cue that, rather than trying to understand that infinitesimal fraction that can be understood, he should just conclude that it was all impossible to understand. And then he thought that maybe that was the Big Secret that everyone was trying but failing to reach: That it’s impossible for anyone to understand any of science.
And then, of course, he told an AI to validate his idea that it was impossible to understand science, and so he shouldn’t try, and that it really is the Big Secret that science isn’t actually understandable.
Nuts.
Well… …Yes.
It’s not that F doesn’t equal M X A it’s that we misinterpret what these are and why it is M X A.
Reality is procedural logic and logic applies across domains.
By thinking symbolically most people come up with the wrong answer to most physics questions. For instance a weight placed on the edge of a frictionless perfectly flat table would oscillate like a pendulum as it fell down toward the center of the table which is closer to the center of the earth. Most laypeople can not properly frame even simple physics and get a correct answer. 50% of aviation engineers and 90% of pilots mistakenly believe a plane can’t take off from a conveyor belt moving toward it! Everybody has every answer but they can’t even do simple math in the real world. Scientists can do it but then they never have come up with a unified field theory.
You are mistaken. AI’s often interact with people procedurally.
Ah, the airplane/conveyer belt story again. I believe that all that does is spin the wheels.
Nobody asked you to surrender.
I don’t believe anything until I can incorporate it into my models. This has forced me to reorganize my models repeatedly to reflect new things I knew to be true. I do not tolerate incongruities in my models.
I estimate odds of things being real based on the premises that reality exists and everyone makes sense. I define “certainty” as being 99.9% sure. I am 99.9% sure Moscow exists. I assume reality exists and people make sense.
It was easy enough to take a test and tell the teacher what they wanted to hear because I made models of what i was taught. These models were dismantled and pieces reused.
To an AI, all a person even is is a generator of strings of symbols.
My father told a story about a man pigeon holed in a looney bin back in the 30’s. After many years he was finally examined by a professional because could sketch complex electrical circuits on a napkin.
Do you know of any cites that show that this actually occurred? Otherwise, it’s just a tall tale.
Could you be a good person and be standing quietly wayyy over there?
Thank you so very, very much.
You could look for one.
He also told a story of meeting a headless chicken.
We’ve all had bad days.
I’m not the one citing that story as an example of anything. Not going to do your homework for you.
I thought your dad only believed half what he saw.
So maybe he knew the dude in a looney bin drawing intelligent circuits or maybe he didn’t.
The math is 50%, on that
Let it trickle down to you knowing not to believe what you hear and telling us not to.
100% of us will not buy this. At. All.
There’s some very simple numbers for ya.
Yeah. That ain’t how this board works.
We may be using terms differently. AI creates tokens and, yes, these are “symbols” but it doesn’t think or manipulate words to process data. What it does is convert words to what can be computed and then translate that output back into words. It’s COGNITION that is symbolic or procedural. Consciousness is procedural but modern humans use a symbolic overlay called the brocas area in order to think. All of us do some procedural thinking but some do very little. AI has to guess the framing of a prompt, process it, and then convert it back into mthe same framing.
I wish we were still talking about USB-C.