AI is ridiculous concept that many misinterpret.

If the human neural network employs electro/chemical signaling and processing and bacteria already use a purely chemical language to communicate (quorum sensing), are we considering adding internal chemical processing to the purely electrical networks of AI?
As I understand it, we are now creating materials which are chemically reactive and able to perform work.
(a slime mold walks by chemical contractions of the cell)

Please Register or Log in to view the hidden image!

Artificial muscle
Description
Artificial muscles are actuators, materials or devices that mimic natural muscle and can reversibly contract, expand, or rotate within one component due to an external stimulus. Wikipedia

 

Log in or Sign up to hide all adverts.

Incorrect, as has been pointed out.

Nope. We do in fact build computer neural networks. They support artificial intelligence applications.

No one is assuming that.

OK. So to you, a computer with a neural network is not a computer. That's fine - but you should be aware that most people use the real definition and not yours.

Of course it can.

Google PUFs.

 

Log in or Sign up to hide all adverts.

I don't think the medium matters if intelligence is an emergent property of its function, not it's material.

 

Log in or Sign up to hide all adverts.

@ Billvon,

I had pretty much given up on this thread when a member of sciforums staff with over 32 000 postings here stated,

Given that the Google Home is not at all intelligent, how do you think it makes sense of your voice commands? Do you have any idea of how difficult a problem it is for a computer to make sense of your voice? That was problem that computer scientists worked on cracking for over 50 years since the first digital computers were invented. Do you think that it's done through a series of if-then statements?

It became obvious to me that some here are a tad slower than others. That statement does not speak well of the type of thinking prevalent here.

Obviously to some, a computer can input voice recordings and break them down into modules of tones/amplitudes assigning numerical values which are then COMPARED using a series of IF/THEN type statements (in any computer language) and then compared to the values of pre-existing words and statements held in memory.

Now.... You wish to challenge the scientific norm by also claiming the impossible.

I had stated in my last post,

A computer cannot even choose a random number.

You responded,

Of course it can. and added Google PUF's.

That would be a HARD NO.

The only way a computer can get numbers is by seeking them as what we might deem random outside input but I had demonstrated that in my last answer.

For a more day-to-day example, the computer could rely on atmospheric noise or simply use the exact time you press keys on your keyboard as a source of unpredictable data, or entropy. For example, your computer might notice that you pressed a key at exactly 0.23423523 seconds after 2 p.m.. Grab enough of the specific times associated with these key presses and you’ll have a source of entropy you can use to generate a “true” random number.
https://www.howtogeek.com/183051/htg-explains-how-computers-generate-random-numbers/

PUF's also are not random as they use numbers grasped from sources other than the program.


This is not my opinion. This is just the way it is. THIS IS A MATH QUESTION ALSO... (how can you use math to generate a random number?)

Mathematician answer: John Von Neumann was aware of this limitation, of middle-squares and randomness generally, andin a 1951 paper offered his famous quote (one of them): "Any one who considers arithmetical methods of producing random digits is, of course, in a state of sin. For, as has been pointed out several times, there is no such thing as a random number—there are only methods to produce random numbers."

See also,
https://en.wikipedia.org/wiki/Hardware_random_number_generator

See also,
https://en.wikipedia.org/wiki/Pseudorandom_number_generator

See also,
NOTE: I was going to cut and paste some information from this site but didn't want to waste my time,
It is however descriptive of various types of random numbers and the terminology being discussed. If you do not reconsider your position and argue that true random number generation by software is possible (without outside sources of input/data collection). All computer processes involved in even pseudo-random number generation use if/then COMPARISONS and algorithms to try to approximate random. This has not been fully developed but hopefully better math will occur for this some day. Perhaps it is impossible if Von Neumann was correct and it is an impossible feat.|
https://courses.physics.illinois.edu/phys466/sp2013/lnotes/random_numbers.html

You could just GOOGLE "Can a computer generate a random number?"
https://www.google.com/search?q=can+a+computer+generate+a+random+number&rlz=1C1AVNG_enCA771CA771&oq=Can+a+computer+generate+a+random+number&aqs=chrome.0.0l3.9959j0j4&sourceid=chrome&ie=UTF-8

and read for hours....


Here is a the very first quote that pops up on my phone...
Even though a computer is absolutely unable to create a random number, it is good at introducing too many variables to predict! Generating truly random numbers in software is indeed not possible as others have pointed out, however it is possible with hardware to build a device which can generate truly random numbers
https://softwareengineering.stackex...it-impossible-to-produce-truly-random-numbers

One such method to get a truly random number (if you read above links)

• Truly Random- numbers exhibiting "true" randomness, such as found by counts of a Geiger measuring radioactive decay.
  

This is akin to the method I stated in my last post. If you wanted a truly random number from 1 to 60 you could look at the seconds timestamp when the process is being initiated. This would give a true random number ONCE. To repeat random number generation you would need a non mathematical approach. You would need outside input.

As the above stated one such method would be measuring radioactive decaying. This would give the program a source of random numbers.

So.. When you state things like

"Of course it can"

to my statement that software cannot generate random numbers.

You seem to be misinformed, unable to google for 1 minute prior to answering, trolling, or flat out wrong.


Your PUF's answer is simply an external method of achieving inputs akin to the geiger counter. I had stated in my last answer that they had chips that are used for random number generation nowadays which is also akin.

But a computer can never "THINK" of a random number. It is impossible now and impossible for the next 10 000+ forever years.

You are a member with 15 000+ postings.

That is why I had abandoned this thread. Too much trolling, arguing for arguments sake (grandstanding, etc).

I SAID IN LAST COMMENT...

A COMPUTER CANNOT EVEN CHOOSE A RANDOM NUMBER. It can be handed one with some extreme examples being a geiger counter measuring radioactive decays. It just cannot ever choose one itself.

Now I do expect arguing for the sake of it. However; I will simply point to the work of others again if such notions are challenged or twisted to the point I feel compelled to respond further..

 

Sure it can.

It can also just listen to those words, then learn what they mean through repetitive training. No IF/THEN statements.

I see you did not google PUF's.

PUF's are physically unclonable functions. A unique random number, right there in the computer. Not based on anything external.

Want lots of random numbers? Use a dynamic PUF. A simple reverse biased transistor or diode feeding into a DC-centered CMOS comparator will give you all the noise you want. This noise can then generate any sort of random number you like - without any external inputs. And that noise will be based on random, internal physical events, like the noise caused by electrons jumping a bandgap.

You are apparently somewhat fact challenged.

You do not need outside input. See above.

Factually incorrect. Many Intel processors, for example, contain a built-in, hardware based random number generator that can generate random numbers rapidly without any external input.

 

To me Ai is very real .

A shame really .

All thought will become electronically based .

Organic thinking , based on life perspective , could be lost forever .

 

No

At the moment I am thinking about the 1,000s of neurons involved in thought. I think the synapse is not leak proof and the leaks form unintended (what else) connections. That brings into question how do you direct your thoughts? to draw on the stuff you are trying to remember?

Anyway back to the feral (leak) thoughts. Again I have spent 37 years on this problem (or maybe about a hour). If the brain detects a pattern already in memory, it shuffles that pattern, originated from feral thoughts, to the storage area. Compare already know to new from feral and melds the two together. This system opens up a new pathway and further options to consider (about the problem being thought about)

As I understand the process a lot of this takes place in the sub conscious

Conscious thought is of the type when you try to remember what you were doing at 10 am yesterday

And I agree none of the neurons handling the process understand meaning, but but but as a collection do understand, via feedback from concessness, I am thinking about what I was doing at 10 am yesterday

Please Register or Log in to view the hidden image!

 

@ Billvon, you stated.
Sure it can.

It can also just listen to those words, then learn what they mean through repetitive training. No IF/THEN statements.

You are the second senior member of this forums (one was staff) that has suggested If/then statements are not necessary for certain AI processes.

I do not wish to get into the long intricacies of this very dumb statement so I will just say that no program that has a subroutine can operated without an "If/then" statement in the computer language of your choice. It also does not "LISTEN" (personifying). It recieves input through microphones that are measured into quantifiable data and at least STORED for further comparisons if not directly compared to words and phrases on hand.

It seems many of the proponents to AI here have little understanding of how prevalent IF/THEN (in the computer language of your choice) statements are.

Computer programming is easy to learn because it is mostly one line of command (If/then statements in computer language of your choice) used repeatedly.

@ BILLVON again,

This entire thread is about software intelligence. You have stated I do not understand your PUF's and do not seem to understand half of what I wrote about externally getting true random numbers.

One such method to get a truly random number (if you read above links)

• Truly Random- numbers exhibiting "true" randomness, such as found by counts of a Geiger measuring radioactive decay.

This is akin to the method I stated in my last post. If you wanted a truly random number from 1 to 60 you could look at the seconds timestamp when the process is being initiated. This would give a true random number ONCE. To repeat random number generation you would need a non mathematical approach. You would need outside input.

As the above stated one such method would be measuring radioactive decaying. This would give the program a source of random numbers.

So.. When you state things like

"Of course it can"

to my statement that software cannot generate random numbers.

You seem to be misinformed, unable to google for 1 minute prior to answering, trolling, or flat out wrong.


Your PUF's answer is simply an external method of achieving inputs akin to the geiger counter. I had stated in my last answer that they had chips that are used for random number generation nowadays which is also akin.

But a computer can never "THINK" of a random number. It is impossible now and impossible for the next 10 000+ forever years.

You are a member with 15 000+ postings.

That is why I had abandoned this thread. Too much trolling, arguing for arguments sake (grandstanding, etc).

I SAID IN LAST COMMENT...

A COMPUTER CANNOT EVEN CHOOSE A RANDOM NUMBER. It can be handed one with some extreme examples being a geiger counter measuring radioactive decays. It just cannot ever choose one itself.

Now I do expect arguing for the sake of it. However; I will simply point to the work of others again if such notions are challenged or twisted to the point I feel compelled to respond further..

So when you say,
I see you did not google PUF's.

PUF's are physically unclonable functions. A unique random number, right there in the computer. Not based on anything external.

Want lots of random numbers? Use a dynamic PUF. A simple reverse biased transistor or diode feeding into a DC-centered CMOS comparator will give you all the noise you want. This noise can then generate any sort of random number you like - without any external inputs. And that noise will be based on random, internal physical events, like the noise caused by electrons jumping a bandgap.

You are lying or attempting strawman argument( look it up). As I had SAID in my last post...
Your PUF's answer is simply an external method of achieving inputs akin to the geiger counter. I had stated in my last answer that they had chips that are used for random number generation nowadays which is also akin.

So to any literate... I had addressed PUF's. It is using computer hardware as basis for some calculations in achieving random numbers.

OH... DID THAT REPLY FROM ME MENTION CHIPS? DID IT? HUH? HUH? HUH? (Im getting tired of repeating myself to people who cannot grasp simple programming is mostly if/then statements in computer language of your choice)

DID I MENTION CHIPS IN THAT SENTENCE.... when I said...
I had stated in my last answer that they had chips that are used for random number generation nowadays which is also akin.

So it almost sounds like you are suggesting I did not vehemently suggest chips and external factors were THE ONLY METHOD a computer can get a truly random number.

Again I said
A COMPUTER CANNOT EVEN CHOOSE A RANDOM NUMBER.

To which Billvon responded,
Factually incorrect. Many Intel processors, for example, contain a built-in, hardware based random number generator that can generate random numbers rapidly without any external input.

HMMMMMM. Could you mean a,
I had stated in my last answer that they had chips that are used for random number generation nowadays which is also akin.

Mr Billvon can imply that getting a random number from an external source (Like looking at a geiger counter is random numbers) is proof of thought in software when it seeks such numbers externally (even a random number chip is external to the software)

But hey... Obfuscate, lie... and say whatever saves you from admitting you were wrong. I would expect nothing less.

They also do not give true random numbers (BTW!), but as we will see James R says we cannot generate random numbers either in our brains (facts only please Mr R.)

Again James R postulates the woo that the brain is the same function of a computer.
By saying
Neither can you [choose a random number]. Would you say that human intelligence is therefore impossible, too?

@ Michael...

I did argue much of what you did on earlier pages. I suggested our creativity could sprout from "poor memory", etc. I.e. Leaks.


I enjoy the argument that perhaps we are only creatures that can input compare, compare, compare, and then output but it remains woo at this point in our history.

A computer can do nothing except input, output and compare. If these guys wish to equate the human brain to the exact same processing... input, compare, output.. then that is indeed interesting and I have even tried to align myself to such thought.. I just reject it for now without actual evidence.

 

That's an assumption that is in question here, is it not?
I'm not saying it is, or it isn't, but I don't want to beg the question.

That's the big question, is it not?
First, though, do you see a difference between the systematic approach and semantic understanding?
If so, then a key question is how the latter can be achieved by the former.
But please don't confuse me for someone who thinks AI is a "ridiculous concept"; I am merely offering an argument that highlights a potential roadblock.
Not necessarily one that can not be overcome, but I'd hope it takes more than just "oooh, complexity, that'll do it!".
I also have no issue with it currently being an "unknown" that might likely be overcome, but that doesn't stop it being an issue to address.

It's the same question I'd have with p-zombies: how can I tell, for example, that you are conscious, that you actually experience things as I do, or at all?
And to me it's among the first questions we need to answer: what is intelligence, being the first, and whether it is something wholly judged by how it appears to other people.
Although whether one can answer the first without addressing the second, I am doubtful.

 

I completely agree that computers are 100% machines.
I'm not sure what's "true thought", though. I will assume it's the kind of thought human beings do. So, sure, computers don't and couldn't do that but a car doesn't do walking either and it's not exactly a problem at all. Brains are like computers essentially physical processes. We don't actually need to replicate the wet tissue of a brain into a computer to make it perform essentially the same rational functions. Current AI machines are very far from coming near the brain in terms of raw computational power but one day maybe they will, and beyond.
We don't need AI machines to becomes the replicates of human beings, with moral values, aching pains, blissful joys and such. But we do need intelligent machines. In fact, we need machines more intelligent than us. And, probably, we will get them. It seems almost a certainty, although I wouldn't bet on the timeline. Current research seems hopeless, but at some point some dude will find the correct angle.
And you're wrong about imagination. There's no impossibility in terms of the principle, only imagination requires computational power and some dude to code the algorithm.
EB

 

The only working definition of intelligence I've come across that makes any sense is the ability to adapt learned things to solving a heretofore un-encountered problem.

Humans can do this.
Humans we regard as very intelligent can do this well.
Advanced animals can do it to some degree.
Computers are starting to be able to do this but only in very specialized fields.

In my view, there is a near 1:1 correlation between what we intuit as "intelligence" and the above definition - meaning I think its a good definition for what we're trying to nail down.

 

The question is do we consider him to be intelligent?

Please Register or Log in to view the hidden image!

 

As with binary systems, Billvon.

 

Humans aren't computers and AI will never replicate human thought. It doesn't need to do that. Humans do things one way and AI could do things another way. What matters is the task and how it gets done (or that it gets done).

Computers are better at some things than humans. Humans are better at some things than a computer is likely to ever be.

All that can be true and there can still be high level AI involved in many tasks.

 

That is absolutely true. Specifically, it is true for neural networks. They do not return yeses or nos, they return weights. Just like your brain does.

At a very low level, of course, computers are binary, just as your neurons are. They fire or they don't. But at higher levels, the outputs are weights, not yes/no decisions.

In other words, just like you do. You receive inputs through sensors (your ears) store it for a short time and then compare it to words you have heard in the past. You then usually get an accurate account of what was said. Sometimes it's not 100% clear and you think "did he say 'pole' or 'pool?' I will go with pool because it makes sense in the sentence."

Again, just like a neural network trying to understand the same word.

And unfortunately you have not realized that in the 30 years since you have studied BASIC computing has evolved.

No. A PUF is internal. It comes from inside the computer. It does not come from outside input. Likewise, more sophisticated systems (like Intel's) can generate large numbers of completely random numbers without external input. You are factually incorrect.

Nope. A PUF does not look to external events like a Geiger counter does. It uses internal events.

It absolutely can. I have given you two examples.

So you have stated two things that are completely incorrect, and you still apparently wish to be taken seriously. Why not educate yourself first, rather than relying on Googling and the BASIC you learned 30 years ago? It will be a better use of your time than posting incorrect nonsense over and over.

 

Yes, exactly!

 

Or another way to put it is that... Due to what should be our desire to not go extinct via the advent of superior rivals, we would not want them to be as free, capable, and open to self-development as human beings. However, since we're not universally devoid of stupidity, what we "ought do" or "should avoid policy-wise" is not necessarily going to become fact.

EDIT: Probably taking too much the sensationalistic outcome there. AIs and robots would become less efficient and reliable if they were designed to be more like us, with personal interests, self-programming capacity, etc conflicting with their intended tasks. Last thing society needs is occasionally unpredictable tools running lose along with the similarly quirky and occasionally unpredictable biological inhabitants.