• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

✳ This is it ... the scientific topic we always wanted to discuss ...

Wes

Major Contributor
Forum Donor
Joined
Dec 5, 2019
Messages
3,843
Likes
3,788
The computer has got to be the most misunderstood invention ever. It is so misunderstood that some people are going around suggesting that our existence is a computer simulation. Movies are made about them. AI is regularly discussed as something uncontrolled having a life of its own. All complete rubbish. In reality, computers are very simple devices in function and those who understand how they work are not impressed with what they are capable of. Computers use transistors to indicate on and off positioning as a storage for the "digital" representation (based on agreed upon coding and protocols) of analog sound or visual end products. They are primitive task masters. Powerful yes, but not magical in any sense. They cannot do anything which cannot be done with simple pen and paper, they just do it incredibly fast. And there is no such thing as AI, there are only computer programs which determine what a machine does. The computer is nothing more than a gopher. If you can simulate life using a computer, you can also simulate a life using pen and paper. You could simulate a life using beans and peas (beans for 1s, peas for 0s) too if you had 11 billion of them and loads of time (energy bars would be helpful too). Actually you could do it with only 1 bean and 1 pea if you really have the time. So there it is, life is a pea and a bean.

You misunderstand AI
 

Wes

Major Contributor
Forum Donor
Joined
Dec 5, 2019
Messages
3,843
Likes
3,788
It really SHOULD be free of charge actually. It is too critical now for anyone to go without based on income. Heck, you can't even get certain things done now without it and it is getting more critical each day.

compare the US to South Korea - we pay out the wazoo for much slower internet
 

lashto

Major Contributor
Forum Donor
Joined
Mar 8, 2019
Messages
1,045
Likes
535
The reality is that we have in germany some kinde of information oligarschie. Nothing gets cheaper.
Well, internet got much cheaper during last 10-20 years. Not much else though. Hard to think of something that got cheaper and does not belong to the newtech/IT category.
 

tomtoo

Major Contributor
Joined
Nov 20, 2019
Messages
3,607
Likes
4,514
Location
Germany
Well, internet got much cheaper during last 10-20 years. Not much else though. Hard to think of something that got cheaper and does not belong to the newtech/IT category.

I talk about the last 5 years. A dsl/phone stagnates at around 20€/month.
 

cathodeb

Member
Joined
Oct 12, 2020
Messages
63
Likes
41
You misunderstand AI

It does nothing for anyone to exclaim "You misunderstand AI", without saying why or how, but I'm going to suppose you already knew that (or maybe I'm giving you too much credit). In any event, I find your posts in general underwhelming so don't confuse this as a request for information.
 

MRC01

Major Contributor
Joined
Feb 5, 2019
Messages
3,424
Likes
4,030
Location
Pacific Northwest
Free will? You think we have?
We live in a probabilistic, non-deterministic universe. That means the notion of freewill is not necessarily against the laws of physics.
That doesn't mean we have freewill. Given the current state of human knowledge, the definition of freewill and whether we have it is an open question.
Personally, I feel that I do have freewill, and this feeling is so compelling, the examples are so convincing, I must believe it is true, until and unless I encounter incontrovertible evidence otherwise.
And I feel this belief is consistent with the scientific method, which is to take our observations as primary evidence and build theories consistent with them.
 

cathodeb

Member
Joined
Oct 12, 2020
Messages
63
Likes
41
.
Personally, I feel that I do have freewill, and this feeling is so compelling, the examples are so convincing, I must believe it is true, until and unless I encounter incontrovertible evidence otherwise.
.

I would go a step further and say that even if you "encounter incontrovertible evidence otherwise" you will STILL believe that you have free will because, as you point out, you feel it and it is compelling - much more compelling that any bit of countervailing logic will ever be. Pirandello had it right - "right you are (if you think so)". The human experience is far more primary than science when all is said and done.
 

Wes

Major Contributor
Forum Donor
Joined
Dec 5, 2019
Messages
3,843
Likes
3,788
It does nothing for anyone to exclaim "You misunderstand AI", without saying why or how, but I'm going to suppose you already knew that (or maybe I'm giving you too much credit). In any event, I find your posts in general underwhelming so don't confuse this as a request for information.

you could go look it up - that will save everyone some time

a science education will help you understand my posts better
 

MRC01

Major Contributor
Joined
Feb 5, 2019
Messages
3,424
Likes
4,030
Location
Pacific Northwest
... [computers] cannot do anything which cannot be done with simple pen and paper, they just do it incredibly fast. And there is no such thing as AI, there are only computer programs which determine what a machine does. ...
That has a thread of truth but it misses an important point. AI (or more specifically, Machine Learning) refers to algorithms that have a few interesting properties:

1. They discover patterns in data, without being "told" what the patterns are. The programmers who wrote the algorithms have no idea what patterns it will find, and the same code finds different patterns when pointed at different data.

2. The algorithms use these patterns to make decisions (typically classifiers or regressors). Their decisions, being based on the patterns, which are based on the data, are inherently unpredictable even to the programmers who wrote the algorithms.

3. Given more data, they make better decisions. This happens "auto-magically" without the programmers intervening, as the algorithms incorporate error correction feedback.

The above 3 characteristics sound similar to the process that we humans call "learning". This doesn't mean that computers learn in the same way humans do, or that computers "understand" anything. Despite the above characteristics, computers are still nothing more than Turing machines as you describe. But it's quite interesting that we can create "mindless" Turing machines that exhibit the above characteristics that at least resemble learning. It means computers can come up with insights that their creators did not envision.
 

raif71

Major Contributor
Joined
Sep 7, 2019
Messages
2,333
Likes
2,535
The computer has got to be the most misunderstood invention ever. It is so misunderstood that some people are going around suggesting that our existence is a computer simulation. Movies are made about them. AI is regularly discussed as something uncontrolled having a life of its own. All complete rubbish. In reality, computers are very simple devices in function and those who understand how they work are not impressed with what they are capable of. Computers use transistors to indicate on and off positioning as a storage for the "digital" representation (based on agreed upon coding and protocols) of analog sound or visual end products. They are primitive task masters. Powerful yes, but not magical in any sense. They cannot do anything which cannot be done with simple pen and paper, they just do it incredibly fast. And there is no such thing as AI, there are only computer programs which determine what a machine does. The computer is nothing more than a gopher. If you can simulate life using a computer, you can also simulate a life using pen and paper. You could simulate a life using beans and peas (beans for 1s, peas for 0s) too if you had 11 billion of them and loads of time (energy bars would be helpful too). Actually you could do it with only 1 bean and 1 pea if you really have the time. So there it is, life is a pea and a bean.
Of course the computer mentioned here is not the same as the computer as we know it. The computer we know now is binary in nature but we are already talking about quantum computer that is more powerful and this is at our "level". We may not know what kind of computer that governs this reality that we are living in. Of course this is just all conjecture but just want to point out that the "computer" that was mentioned is not the kind of computer that we are working with now, that's for sure. :)
 

MRC01

Major Contributor
Joined
Feb 5, 2019
Messages
3,424
Likes
4,030
Location
Pacific Northwest
... The computer we know now is binary in nature but we are already talking about quantum computer that is more powerful and this is at our "level". ... just want to point out that the "computer" that was mentioned is not the kind of computer that we are working with now, that's for sure. :)
As far as we know, Quantum computers can't solve NP-hard problems (like the traveling salesman) in polynomial time, or any faster than classical computers. There is another complexity class (suspected to be) between P and NP, called BQP, for "Bounded Error Quantum Polynomial Time". Quantum computers can solve problems in BQP faster than classical computers. But the pragmatic value of this is not fully known, and debatable. There is only 1 practical problem I know of in BQP, deriving the prime factors of large integers (Shor's algorithm). So (we suspect) that's 1 thing that a quantum computer can do faster than a classical computer. Even there, pragmatically speaking, classical computers are still faster because nobody's yet devised a quantum computer big enough to leverage this theoretical advantage on big enough numbers to make a difference.

PS: looks like there are a few other practical suspected-BQP problems: https://arxiv.org/abs/quant-ph/0606179

Either way, a quantum computer is still just a machine that follows formal logic / code / algorithms. The reason it can be faster than classical computers with some kinds of problems is because it operates probabilistically leveraging the fact that a qubit can be in multiple states simultaneously. But this doesn't change what it can do, it only changes its method and efficiency. Whether it is "at our level" remains to be seen. We don't know what our own minds really are, or how they work.
 

tomtoo

Major Contributor
Joined
Nov 20, 2019
Messages
3,607
Likes
4,514
Location
Germany
Of course the computer mentioned here is not the same as the computer as we know it. The computer we know now is binary in nature but we are already talking about quantum computer that is more powerful and this is at our "level". We may not know what kind of computer that governs this reality that we are living in. Of course this is just all conjecture but just want to point out that the "computer" that was mentioned is not the kind of computer that we are working with now, that's for sure. :)

Thats not the point. You can simulate quantum computers on digital machines. It takes just longer. So the way any imagined machine works is not realy from interest. If we would live in a simulation? What would change? Maybe we are all just thoughts in a bolzmanbrain? What would change?
 

tomtoo

Major Contributor
Joined
Nov 20, 2019
Messages
3,607
Likes
4,514
Location
Germany
We live in a probabilistic, non-deterministic universe. That means the notion of freewill is not necessarily against the laws of physics.
That doesn't mean we have freewill. Given the current state of human knowledge, the definition of freewill and whether we have it is an open question.
Personally, I feel that I do have freewill, and this feeling is so compelling, the examples are so convincing, I must believe it is true, until and unless I encounter incontrovertible evidence otherwise.
And I feel this belief is consistent with the scientific method, which is to take our observations as primary evidence and build theories consistent with them.

There we end. Whats the definition of free will?
 

cathodeb

Member
Joined
Oct 12, 2020
Messages
63
Likes
41
1. They discover patterns in data, without being "told" what the patterns are. The programmers who wrote the algorithms have no idea what patterns it will find, and the same code finds different patterns when pointed at different data.

I don't believe that computers "discover" anything. They receive data and based on a program, "do" something or not. PERIOD. People just don't understand how simple this is. Computers cannot do ANYTHING that they were not programmed to do. No matter how complex a computer's actions (think of a computer that is robotic which can buy a ticket to Alabama, join a rock group and become the boss of a mafia family, for example). This fancy computer owes all its success to its programmer and in fact is as unconscious and as able as a carpet. And it changes nothing that the programmer never imagined the mafia angle. A computer is nothing more than a bean and pea counter and it will do exactly what the programmer tells it to do and nothing more or less regardless of whether the programmer knows the outcome based on unknowable data.

2. The algorithms use these patterns to make decisions (typically classifiers or regressors). Their decisions, being based on the patterns, which are based on the data, are inherently unpredictable even to the programmers who wrote the algorithms.


It is not the "decisions" which are unpredictable, it is the OUTCOMES. The decisions are by definition predictable because they are made BASED ON CODE. If you want to press the point further, perhaps you could give a concrete example of a computer doing something unpredictable and if you could, please explain HOW IT DECIDED TO MAKE that decision with referring to code. Even sloppily written code with unintended consequences cannot render a computer unpredictable, exciting though the possibility sounds. A computer is not a magical thing, it is a simple thing and all it does it push 1s and 0s around based on the program that is executing.

3. Given more data, they make better decisions. This happens "auto-magically" without the programmers intervening, as the algorithms incorporate error correction feedback.

Computers can learn and "make better decisions" BASED ON THE LOGIC I EMPLOY ONLY. The computer itself only pushes 1s and 0s around and is as dumb as particle board. Learning is not auto-magical. The computer is the exact opposite of "magical". Even a hammer is more magical, and by far.

The above 3 characteristics sound similar to the process that we humans call "learning"..

I never said that computers can't learn. I'm simply saying that a computer can only do what it is programmed to do. If I tell program it to learn, it will learn and it will learn exactly the way I programmed it to learn. It doesn't have a mind of its own. It is not intelligent because it is not an IT. It cannot do WHAT IT WANTS TO DO because there is no IT. The fetishization of the computer is really mind boggling to me.
 
Last edited:

cathodeb

Member
Joined
Oct 12, 2020
Messages
63
Likes
41
but just want to point out that the "computer" that was mentioned is not the kind of computer that we are working with now, that's for sure. :)

newer computing methods including quantum computing have to do with the speed of processing (HOW the computer achieves what it does), not the nature of what the computer does.
 

StefaanE

Addicted to Fun and Learning
Forum Donor
Joined
May 12, 2020
Messages
528
Likes
929
Location
Harlange, Luxembourg
AI lack what cockroaches have, and that is motivation. I am amazed that people seem conflate intelligence with the drive to behave in such a way that procreation happens. Without artificial motivation, artificial intelligence sits on its butt.
 

Wes

Major Contributor
Forum Donor
Joined
Dec 5, 2019
Messages
3,843
Likes
3,788
I don't believe that humans "discover" anything. They receive info and based on genetics, "do" something or not. PERIOD. People just don't understand how simple this is. Humans cannot do ANYTHING that they were not programmed to do. No matter how complex a human's actions (think of humans which can buy a ticket to Alabama, join a rock group and become the boss of a mafia family, for example). This fancy human owes all its success to its genetics and in fact is as unconscious and as able as a carpet. And it changes nothing that the genetics never imagined the mafia angle. A human is nothing more than a bean and pea counter and it will do exactly what the genetics tells it to do and nothing more or less regardless of whether the genetics knows the outcome based on unknowable data.


Humans can learn. I'm simply saying that a human can only do what it is programmed to do, by genetics and developmental "switches" and its learning and experience. If I tell human to learn, it will learn and it will learn exactly the way genetics programmed it to learn. A Human will claim to have a mind of its own, but don't. A Human cannot do WHAT IT WANTS TO DO because there is no free will. The fetishization of the Human is really mind boggling to me.
 

raif71

Major Contributor
Joined
Sep 7, 2019
Messages
2,333
Likes
2,535
There we end. Whats the definition of free will?
I'd like to say that we can make plans and act upon those plans. That is free will. For example I plan to buy a dac/amp. After that I browse around ASR and found what I like. I make a trip to the local audio store and bought said dac/amp. I was responsible for that purchase and I made a choice or at the very least I believed I made a choice. :)
 
Last edited:
Top Bottom