AI Doesn't Exist

[Glitch]

What the uninformed - such as the media - call Artificial Intelligence is not actually Artificial Intelligence. They just don't know any better.

Artificial Intelligence (AI) is the ability of a computer to go beyond its original programming and solve problems that it was never specifically programmed to accomplish. A computer that follows its programming exactly as written cannot be construed as Artificially Intelligent, no matter how clever its programming may be. Those systems are called "Expert Systems" and they are not AI.

We have Expert Systems capable of making accurate medical diagnosis. We have Expert Systems capable of flying aircraft and spacecraft. We have Expert Systems doing a wide variety of complex tasks, and none of them are AI. Expert Systems are when the programmer(s) write their expertise into a form that the computer can calculate. It is therefore not the expertise of the computer, but the expertise of the programmer(s) involved.

A computer can only be considered artificially intelligent when it can exceed its own programming by performing tasks it was never specifically programmed to accomplish. If a program that was designed and developed to diagnose medical issues, and it also provided a solution for cleaning up an oil spill, for example, then that would be an example of true AI.

Unfortunately, what we have are a bunch of ignorant fools in the media calling everything AI, when none of it is. Everything the media calls AI is actually an Expert System and not AI at all.

As someone who has worked in the field since the 1970s, AI is quite impossible to achieve with binary computers. It may be possible with quantum computers, but that is at least another 50 to 100 years into the future before that will become a reality.

 

[StarvedForChange]

What the uninformed - such as the media - call Artificial Intelligence is not actually Artificial Intelligence. They just don't know any better.

Artificial Intelligence (AI) is the ability of a computer to go beyond its original programming and solve problems that it was never specifically programmed to accomplish. A computer that follows its programming exactly as written cannot be construed as Artificially Intelligent, no matter how clever its programming may be. Those systems are called "Expert Systems" and they are not AI.

We have Expert Systems capable of making accurate medical diagnosis. We have Expert Systems capable of flying aircraft and spacecraft. We have Expert Systems doing a wide variety of complex tasks, and none of them are AI. Expert Systems are when the programmer(s) write their expertise into a form that the computer can calculate. It is therefore not the expertise of the computer, but the expertise of the programmer(s) involved.

A computer can only be considered artificially intelligent when it can exceed its own programming by performing tasks it was never specifically programmed to accomplish. If a program that was designed and developed to diagnose medical issues, and it also provided a solution for cleaning up an oil spill, for example, then that would be an example of true AI.

Unfortunately, what we have are a bunch of ignorant fools in the media calling everything AI, when none of it is. Everything the media calls AI is actually an Expert System and not AI at all.

As someone who has worked in the field since the 1970s, AI is quite impossible to achieve with binary computers. It may be possible with quantum computers, but that is at least another 50 to 100 years into the future before that will become a reality.

According to britannica.com, Artificial Intelligence/AI is defined as:

Artificial intelligence (AI), the ability of a digital computer or computer-controlled robot to perform tasks commonly associated with intelligent beings. The term is frequently applied to the project of developing systems endowed with the intellectual processes characteristic of humans, such as the ability to reason, discover meaning, generalize, or learn from past experience.

Artificial intelligence (AI) | Definition, Examples, Types, Applications, Companies, & Facts | Britannica

Artificial intelligence (AI), the ability of a digital computer or computer-controlled robot to perform tasks commonly associated with intelligent beings. The term is frequently applied to the project of developing systems with the ability to reason, discover meaning, generalize, or learn from...

www.britannica.com

So, basically, imitate humanity. Your definition seems more like the fictional, "Skynet" from "Terminator" lore.

 

[Bok Tuklo]

I wonder if people can go beyond their programming.

 

[Glitch]

According to britannica.com, Artificial Intelligence/AI is defined as:

Artificial intelligence (AI) | Definition, Examples, Types, Applications, Companies, & Facts | Britannica

Artificial intelligence (AI), the ability of a digital computer or computer-controlled robot to perform tasks commonly associated with intelligent beings. The term is frequently applied to the project of developing systems with the ability to reason, discover meaning, generalize, or learn from...

www.britannica.com

So, basically, imitate humanity. Your definition seems more like the fictional, "Skynet" from "Terminator" lore.

That is a rather silly definition. Considering every program written is "the ability of a digital computer or computer-controlled robot to perform tasks commonly associated with intelligent beings." If all the computer/robot was doing was following the code that was written by a human programmer, where does the computer intelligence come into play?

Following instructions is precisely what computers were designed and built to do. The ability to following instructions exactly as they are written by a human is not an example of intelligence. It might be an example of the human programmer's intelligence, but it certainly isn't an example of computer intelligence.

True AI is indeed fictional. No computer has the ability to do anything beyond its programming. Which means that AI does not exist.

 

[beefheart]

What the uninformed - such as the media - call Artificial Intelligence is not actually Artificial Intelligence. They just don't know any better.

Artificial Intelligence (AI) is the ability of a computer to go beyond its original programming and solve problems that it was never specifically programmed to accomplish. A computer that follows its programming exactly as written cannot be construed as Artificially Intelligent, no matter how clever its programming may be. Those systems are called "Expert Systems" and they are not AI.

We have Expert Systems capable of making accurate medical diagnosis. We have Expert Systems capable of flying aircraft and spacecraft. We have Expert Systems doing a wide variety of complex tasks, and none of them are AI. Expert Systems are when the programmer(s) write their expertise into a form that the computer can calculate. It is therefore not the expertise of the computer, but the expertise of the programmer(s) involved.

A computer can only be considered artificially intelligent when it can exceed its own programming by performing tasks it was never specifically programmed to accomplish. If a program that was designed and developed to diagnose medical issues, and it also provided a solution for cleaning up an oil spill, for example, then that would be an example of true AI.

Unfortunately, what we have are a bunch of ignorant fools in the media calling everything AI, when none of it is. Everything the media calls AI is actually an Expert System and not AI at all.

As someone who has worked in the field since the 1970s, AI is quite impossible to achieve with binary computers. It may be possible with quantum computers, but that is at least another 50 to 100 years into the future before that will become a reality.

Your post doesn't exist.

 

[Glitch]

I wonder if people can go beyond their programming.

Absolutely.

That is how we end up with people like Copernicus, Newton, and Einstein. They questioned what they were taught, and evolved beyond their teachings to reveal new truths that we hadn't considered before.

 

[Slartibartfast]

I am not so sure I agree with this definition. A LLM, when grows to a certain level of complexity, in terms of the number of parameters, has been found to do things that were not originally intended.

One could argue that the LLMs are pattern matching and with a bigger corpus of data, there are more patterns to match, but one could argue just as effectively that a model trained to help write Java code learning how to translate French into English is not the intent of the designer.

I think both arguments are correct at the same time.

 

[Slartibartfast]

I wonder if people can go beyond their programming.

People are just a set of internally competing physical and emotional drives with a layer of consciousness on top. We will never be more than that.

 

[Bodi]

Absolutely.

That is how we end up with people like Copernicus, Newton, and Einstein. They questioned what they were taught, and evolved beyond their teachings to reveal new truths that we hadn't considered before.

They did not evolve, they were programed to be genius's.

 

[Real blank]

People are aware of this.

 

[Rexedgar]

They did not evolve, they were programed to be genius's.

Aren’t/weren’t you an educator?

 

[Glitch]

I am not so sure I agree with this definition. A LLM, when grows to a certain level of complexity, in terms of the number of parameters, has been found to do things that were not originally intended.

A common phenomena among computer developers. Nor are they limited to large language models either. They are also called "bugs" or "features" depending on whether they are a hindrance or useful. That is not an indication of intelligence. That is an indication of a design and/or programming mistake.

One could argue that the LLMs are pattern matching and with a bigger corpus of data, there are more patterns to match, but one could argue just as effectively that a model trained to help write Java code learning how to translate French into English is not the intent of the designer.

I think both arguments are correct at the same time.

It depends on how the program was written, but in either case it still would not be intelligence determined by the computer, but rather by the programmer(s).

During my final exam at the University of Minnesota I had to write a Pascal compiler in Pascal. It was not capable of translating any other language into assembly code, only Pascal. If that program were somehow able to translate C# into assembly code, without any outside interference, then that would be true AI.

You are not going to be able to find any program written today that was developed to translate Java into having the ability to translate French into English. The grammar and syntax rules are not even remotely the same. Maybe in another 50+ years when quantum computers replace binary computers, but not today.

 

[Slartibartfast]

A common phenomena among computer developers. Nor are they limited to large language models either. They are also called "bugs" or "features" depending on whether they are a hindrance or useful. That is not an indication of intelligence. That is an indication of a design and/or programming mistake.

The flaw in your analysis here is conflating mass pattern matching with someone not reading the right post in stackoverflow.

Not the same thing at all.

It depends on how the program was written, but in either case it still would not be intelligence determined by the computer, but rather by the programmer(s).

During my final exam at the University of Minnesota I had to write a Pascal compiler in Pascal. It was not capable of translating any other language into assembly code, only Pascal. If that program were somehow able to translate C# into assembly code, without any outside interference, then that would be true AI.

That depends on the methodology you used.
If you had a model that self trained on pre and post compiled data to build your compiler and it picked up a few languages besides pascal, then yes, it would be emergent behavior and not simply a happy accident.

You are not going to be able to find any program written today that was developed to translate Java into having the ability to translate French into English. The grammar and syntax rules are not even remotely the same. Maybe in another 50+ years when quantum computers replace binary computers, but not today.

Here’s one. It’s quite popular and it can do both quite easily, but not flawlessly.

https://openai.com/index/chatgpt/

My suggestion to you is that you update the framework of your analysis by a few decades. This isn’t simple programs on a PDP 11 anymore.

These programs are giant nth dimensional matrix equations that turn languages and concepts into numerical representations and then run competing algorithms in that space to define and output the “best” answer. What you are describing is deterministic programs with some bits of branching logic, this is not that at all.

This is what you should study. (It’s out of date, but simpler to understand) https://en.m.wikipedia.org/wiki/Word2vec

 

[Glitch]

They did not evolve, they were programed to be genius's.

They were programmed like everyone else during their day. They exceeded their programming to become something more. Which is why we recognize them as genii.

 

[StarvedForChange]

That is a rather silly definition. Considering every program written is "the ability of a digital computer or computer-controlled robot to perform tasks commonly associated with intelligent beings." If all the computer/robot was doing was following the code that was written by a human programmer, where does the computer intelligence come into play?

Following instructions is precisely what computers were designed and built to do. The ability to following instructions exactly as they are written by a human is not an example of intelligence. It might be an example of the human programmer's intelligence, but it certainly isn't an example of computer intelligence.

True AI is indeed fictional. No computer has the ability to do anything beyond its programming. Which means that AI does not exist.

I guess: Argue with Britannica and tell me why I should accept your definition over theirs.

 

[CaughtInThe]

01001001 00100000 01101011 01101001 01101110 01100100 00100000 01101111 01100110 00100000 01100001 01100111 01110010 01100101 01100101

 

[Glitch]

The flaw in your analysis here is conflating mass pattern matching with someone not reading the right post in stackoverflow.

Not the same thing at all.

If you are referring to pattern matching, who determined the pattern for the computer to match? It certainly wasn't the computer. So where exactly does the credit for this pattern matching intelligence belong?

That depends on the methodology you used.
If you had a model that self trained on pre and post compiled data to build your compiler and it picked up a few languages besides pascal, then yes, it would be emergent behavior and not simply a happy accident.

Here’s one. It’s quite popular and it can do both quite easily, but not flawlessly.

https://openai.com/index/chatgpt/

My suggestion to you is that you update the framework of your analysis by a few decades. This isn’t simple programs on a PDP 11 anymore.

These programs are giant nth dimensional matrix equations that turn languages and concepts into numerical representations and then run competing algorithms in that space to define and output the “best” answer. What you are describing is deterministic programs with some bits of branching logic, this is not that at all.

This is what you should study. https://en.m.wikipedia.org/wiki/Word2vec

If I developed the program to accommodate more than just Pascal, then I am coding my own intelligence into the program. The computer isn't thinking outside of the code I provided, and therefore demonstrates no intelligence whatsoever. Simply the ability to follow instructions, as it was originally programmed to do.

I am well aware of how complex equations can become. It is also not a surprise that these programs can provide results that we did not expect. If you have worked with the N-Body problem long enough you can see some surprising results that you might not have anticipated. That does not make them intelligent. They are just better at solving more complex issues at a faster rate than humans. So they can include things like the gravity of objects that we might not have taken into consideration as mere humans, if we program them to take into consideration every object.

Computers are able to process a great deal more data than we can. That in itself could produce results we did not expect. That does not make the computer intelligent. Just better at processing data.

Wikipedia is not a credible source.

 

[Glitch]

I guess: Argue with Britannica and tell me why I should accept your definition over theirs.

Because, as I already explained, their definition covers ALL programs. Whereas my definition only applies to Artificial Intelligence.

 

[Bok Tuklo]

Absolutely.

That is how we end up with people like Copernicus, Newton, and Einstein. They questioned what they were taught, and evolved beyond their teachings to reveal new truths that we hadn't considered before.

How do you know that questioning wasn't their programming?

 

[Glitch]

How do you know that questioning wasn't their programming?

Because there would be a lot more just like them if it had been.

 

[ataraxia]

I wonder if people can go beyond their programming.

Not conservative ones at least.

 

[eman623]

Alan Turin proves it is mathematically possible to write a program to perform any given task, but it's is not possible to create a absolutely general purpose program that can perform any possible task. In that respect, no deterministic AI system can ever become more than it's programming.

Note the key word is "deterministic". What if a sufficiently generalized program could be hooked up to a random number generator to add new random code to its programming? Sort of like the million monkeys typing on a million typewriters and eventually getting Hamlet cranked out. In the vast majority of cases the new programming would not be new or useful or even work and would be discarded. But in some rare instances it could turn out to be beneficial and thus added to the program's repertoire of abilities.

If this sounds like random mutations and evolution, it is not unlike that.

The trick is in deciding what is beneficial. In the real world, organisms don't get to decide for themselves if their mutations are beneficial. Natural selection performs that task. I'm not sure how that process would work in an AI system. Maybe people can be the analogous natural selection mechanism deciding what is beneficial and what is not. Or maybe the terminators will handle that once the AI has randomly nuked humanity out of existence.

 

[soap box]

Ok, from now on we call it ES.

But we'll be the only ones.

 

[Glitch]

Alan Turin proves it is mathematically possible to write a program to perform any given task, but it's is not possible to create a absolutely general purpose program that can perform any possible task. In that respect, no deterministic AI system can ever become more than it's programming.

Note the key word is "deterministic". What if a sufficiently generalized program could be hooked up to a random number generator to add new random code to its programming? Sort of like the million monkeys typing on a million typewriters and eventually getting Hamlet cranked out. In the vast majority of cases the new programming would not be new or useful or even work and would be discarded. But in some rare instances it could turn out to be beneficial and thus added to the program's repertoire of abilities.

If this sounds like random mutations and evolution, it is not unlike that.

The trick is in deciding what is beneficial. In the real world, organisms don't get to decide for themselves if their mutations are beneficial. Natural selection performs that task. I'm not sure how that process would work in an AI system. Maybe people can be the analogous natural selection mechanism deciding what is beneficial and what is not. Or maybe the terminators will handle that once the AI has randomly nuked humanity out of existence.

You raise several good points. Mimicking natural selection through pseudo-random means is one approach.

It may also be possible to develop a program that can improve upon its own code. For example, if an application was developed to analyze code and look for the most efficient means in which to write that code, what would prevent it from analyzing its own code?

As you said, the trick is determining what is beneficial. In my example, it would be the original programmer determining what was more efficient and therefore better, not the computer making that decision. However, the programmer could also be general in their specification as to which was more efficient. Such as, for example, less code means more efficiency. Then the program would look for ways to write the same algorithm with as little code as possible. The intelligence would still lie with the programmer and not the program, regardless of its efficiency.

 

[Bok Tuklo]

Because there would be a lot more just like them if it had been.

There probably are.

Genius needs opportunity. Otherwise it’s the clever child who could not dodge bullets or bombs.