Hacker News Reader

Comment by Psyladine | Hacker News Reader

>since the machine can parse the text faster than you

You had me up until there; the machine doesn't know jack about text. It knows arrays and sequences of numbers according to the rules we've defined them by, for it.

Yuo cna reda tihs raedliy btu teh copmtuer cna't. Your brain is trained by billions of years of evolution for symbolic parsing and pattern pairing, and language is just one flex of that muscle.

Where computers thrive is where we've done the hard work to break down the syllabic system that is inherent to our biology into mathematical abstractions that can be computed by addition. Computers are great at solving problems we've already done, and repeating the steps, nothing more.

Our machines are beautiful, well designed levers. But they don't move anything, they leverage our movement.

View on HN

>Yuo cna reda tihs raedliy btu teh copmtuer cna't.

Why can't it? Isn't that basically what current AI research is doing? Using massively parallel systems to make quick inferences based on existing data sets?

>the machine doesn't know jack about text. It knows arrays and sequences of numbers according to the rules we've defined them by, for it.

If you want to be pedantic and define a computer as the hardware only, sure. The operating system (which contains tools that can in fact parse text) is an essential component in the vast majority of computers in existence. So when I'm discussing computers as a complete, usable unit, then yes, they parse text.

>Our machines are beautiful, well designed levers. But they don't move anything, they leverage our movement.

Well said.

>Yuo cna reda tihs raedliy btu teh copmtuer cna't.

It would be interesting to see what GPT3 would do with this statement.

Edit; just tried it. GPT-3 chatbot understood it instantly.

Cutting edge language processing algorithms trained by professionals in a herding manner for years ought to produce something that solves what is basically a parlor trick of language. It might even be a cheat-around on encoded rules but I don't have the paper.

Still I'm willing to go out on an increasingly thin limb and say GPT3 is still playing a really big game of mad libs, and its 'understanding' of the language is more akin to the extremely rudimentary mathematical factoring of sentence structure and grammatical rules. That is, it could be trained to recycle on what others have done for poetry, and spit out something that sounds like what we've read that we call poetry, but it isn't putting poetry out, because it doesn't know what that means...not in like a conscientious way, but a literal, it has no "idea" which to express through a medium, it's just transcribing according to rules, not from say, first principles which then are interpreted through rules.

Again, too much to demand of a really fast adding machine anyway.

Your view of computers seems overly simplistic. The "adding machine" is just one component in a larger series of abstractions that enable functionality the adding machine is normally too inflexible to perform.

This is because math is a man-made thinking tool, similar to language. It works because we all agree to follow the rules of it, not because those rules are set in stone. The universe operates on its own time, by its own rules. Our constructs are also bound by those rules. You are alive because your entirety is worth more than the sum of your parts, and computers operate on a similar principle. When you shove a wooden board under a rock to pry it loose from the soil, you're leveraging the same forces that allowed you to even wake up that morning, and so does your computer when you jiggle the mouse to wake it.

In my opinion the miracle of consciousness is not the material it gets scribbled onto, but instead the fact it can exist at all.

Reply on news.ycombinator.com