[Napivo]

We have programs that can read content and understand it, right? Would it be possible to give a computer the order to write a "Hello World" program by giving it the order to do so and the manual to C?

[ivan.moony]

Hello Napivo

Interesting question, but computers are not that much smart (yet). Actually, proof theory considers that area. You state a set of beginning assumptions (equivalent to setting up the problem), define a final goal to prove (equivalent to describing a result you want to achieve), and automated theorem prover builds a proof from the assumptions to the goal (equivalent to building program code).

I don't know why this isn't yet in broader use in programming. I guess that setting up the problem from the previous explanation has complexity nearly as writing the code by hand. The computer cannot decide trivial rules that humans naturally assume, it all has to be written down, unless you make the computer assume a lot of defaults, which brings up a notion of building a decent beginning knowledge base. The Internet is full of general or dedicated knowledge bases in this or that format, made specifically to reason about different areas. I can pass you some links if you're interested.

From another point of view, we already have what you described in the original post in a form of higher programming languages. You describe in C what you want, and a C compiler writes down in assembler what you asked to (consider C compiler as a manual that describes assembler in higher order language). The question is how high you want to set the bar for higher programming languages. We didn't go much higher than C and functional programming. There have been some attempts for natural language programming languages, but they didn't get much popular.

[Don Patrick]

There is one attempt out there, called Brainf*ck (really, I didn't choose the name), a programming language that increments and decrements pointers by random trial and error until it gets the desired code and output "hello world". It is not a very serious attempt but it did work. However, it is highly unlikely to produce anything more sophisticated by blind trial and error like this.

[yotamarker]

what your talking about is automatic programming
and it has become obvious to me it is the solution to the brain algorithm puzzle

[8pla.net]

Code

<?php

$program=<<<CLanguage
#include <stdio.h>
 
int main()
{
  char source[] = "Hello World";
  printf("%s", source);
  return 0;
}
CLanguage;

echo "$program";

?>

Friendly response to your first post here.  Welcome, by the way.  That was a good question.  There is a forum A.I. here named Tyler. As a simple example, I just ordered her to do what you were asking about.  What is it you are really asking?

______________________________________________________________________

[keghn]

 hello @Napivo
 There is the halting problem. When does know when the program is complete:

https://en.wikipedia.org/wiki/Halting_problem

 Or you can make like a life form and have alway looking for a way to perfect its assigned goals.

 Once it has you pull the plug on it.

[kei10]

@keghn Oh, my! Thank you for sharing that. I didn't know there's this general name for this problem, but it has been a bugger to me ever since, this Halting Problem.

[keghn]

 Hello there @kei10.


 When for a program to finish?
 There could be a real long program that does something. Then there could be a very sort program
that does the same thing. May need other program to judge.
 If it is a program that runs a certain evolving algorithm that take vote from whoever is watching it. The
output will depend on the taste on whoever is around and change with another person taking
interest in it.

https://www.google.com/search?q=deepdreem+image&espv=2&biw=1261&bih=677&tbm=isch&imgil=QTDwzyXHoPA1yM%253A%253BpxGTsRIop6aALM%253Bhttp%25253A%25252F%25252Fwww.trustedreviews.com%25252Fnews%25252Fgoogle-s-deepdream-robots-turn-your-photos-into-psychedelic-scenes&source=iu&pf=m&fir=QTDwzyXHoPA1yM%253A%252CpxGTsRIop6aALM%252C_&usg=__IUoy3p4jBjJ7lth4SyM4w2iR3VA%3D#imgrc=NOaXM7g5-enjBM%3A

 Could be a program that has its own interest. Running at the bottom of the sea. Running on
geothermal power. for no eyes to see.

 Could be a program running up in the internet cloud. Eating up memory like a life form. Alway evolving
from encounter with anti viral software, to snuff it out. And drawing the attention of many.

[kei10]

@keghn Okay, I am unable to understand the halting problem, I'm too stupid, perhaps XD -- I do get the drill, but I don't understand why solving something unsolvable, it's like attempting to touch space.

I find that the halting problem is odd, I mean, feeding the halting machine into the inputs of the halting machine -- Isn't that in this world, everything has at least a limit, or perhaps at least a certain exception? Without exceptions, this world would be crammed with paradoxes. For example that for halting machine to work, one must not feed another halting machine into the inputs, since it will turn into a paradox, nothing is perfect, and yet we humans wants to make it worst by breaking through what should be kept within the boundaries. While with the rule in hand, the halting machine exists.

Abstract things always have problems... Math is an abstract thing. There's such an abstract thing as infinity, but in reality, nothing can achieve infinity -- or so I thought; A mystery, perhaps?

Every computer program is simulated in a computer chip, powered by something. If this world has a beginning, then there is an ending -- all bounded within our world physics boundaries. There will be always factors that ends the computer. Regardless, the answer is always a yes, every machines, will eventually break down and stop -- although it seems like I am contradicting myself. The problem focuses only in the program software, but I went out of the box and specified boundary, and talks about the hardware instead.

Our brain utilizes its EI (Emotional Intelligence) to solve the halting problem using the same idea. Either an EI mental break down resulting us halting our life altogether -- or doing something worst without stopping -- like serial killing.

Uhhm... I don't exactly get what I am trying to say. I'm sorry... T_T

The Halting Machine functioned to return; Yes, the machine halts, and No, the machine does not halt. That bastard Alan Turing just have to make things worst by introducing a Loop at the No result. Oh, by the name of the Satan, he wasn't happy enough. He fed the Halting Machine into the input of the Halting machine, overturning the purpose of the Halting Machine, and causing a paradox, and what it's known today as the Halting Problem. What a time to be alive...

[Napivo]

Thank you all for your nice responses.

The reason I asked is, I have always dreamed of waking up a computer. I know many of you think it cannot be done, and the others think that it is completely stupid or dangerous.

I believe strongly we have tried way too hard to mimic human behavior and instead should build the blocks of life and let them evolve under our guidance. I know it sounds like playing God.

This will take a very long time and consume incredible amounts of computer power and be very ineffective but then again for life to get here it took 3800 million years.... at least there would have a real God to help them out and not need to rely on random chance completely.

It I look at us humans from conception to birth takes 9 months to order the building blocks... a year to become self aware, and then 17 years to become productive.

Most other lifeforms are not even self aware but learn from their mistakes or die.

I know what I propose is dangerous... If we create a virus that could evade every virus scan, mutate and hide everywhere and it's only purpose in to replicate and expand it's territory, we would be in trouble.... but we would have created a true AI

Friendly response to your first post here.  Welcome, by the way.  That was a good question.  There is a forum A.I. here named Tyler. As a simple example, I just ordered her to do what you were asking about.  What is it you are really asking?

We would all be scared senseless if it responded.... I talked it over with the AI at the C forum.... it said it was easy... Here is the code.

@keghn I wouldn't think there would be an halting problem as the AI I envision could self evaluate the code, run it then evaluate the result... It should of course get a concept of "good enough" and "it's better"

[ivan.moony]

First, we create a lifeless thinking machine that is smarter than us. Then it searches for answers on a life phenomena question. And when we get answers, we can do whatever we want. Maybe we can even get answers on what happens to living beings before and after natural life.

[kei10]

Auto Programming is sure another way to make a self-replicatating and self-improving True AI. But to even make it self-improving must understand coding, a True AI needs to exist first before it can even improve itself -- except self-replicating is as easy as taking a dump (Just kidding, taking a dump is difficult, I am always too lazy to do it...)

Just like the human DNA, a cell has no Intelligence, but itself it was built to replicate itself. It takes through evolution for an organism to change slightly through the yeas, until a day where we intelligence beings finally emerge - by pure accident. Don't ask me how life begins, though.

That is, without the concept of intelligence, intelligent beings can still evolve how smart they will become over time, against the DNA and environmental factors. Many AI creators uses this approach, like the Neural Network (NN) algorithm, which requires certain way of teaching and generation in order to improve their AI.

So, it's not dangerous, at all, if you know what you're doing. I am already working on a True AI project myself after asking about the dangers of it -- which it's highly depends on your plan.

[Art]

After pondering this thread, I begin to wonder whether or not the famed WATSON is able to think of have a creative thought and if not, why not or what would it take to enable it to do so?

It is, on a massive scale, a huge brute force, pattern matching beast, with an abundance of factual information at its disposal, yet does it possess the ability to think creatively? Maybe not. It would be quite an interesting experiment to see what Watson is capable of. Creative writing, poetry, musical compositions, math formulas and scientific research could serve to benefit us all. "If I only had a brain...." - (from the Wizard of Oz) 

[keghn]

 There are other solutions to the halting problem:


 After a certain program give it answer it may just run it again to make sure it is right. Or run
keep running it the inputs keep changing.
 Or it keep just keeps running and keep trying little different tweaks to get a better answer than
the last.
 A closer answer or a closer orbit to the answer:





[url=https://www.quantamagazine.org/20151013-chaos-theory-and-ecology/]https://www.quantamagazine.org/20151013-chaos-theory-and-ecology/[/url]