AI is ridiculous concept that many misinterpret.

Discussion in 'Intelligence & Machines' started by Bob-a-builder, Jun 15, 2019.

  1. arfa brane call me arf Valued Senior Member

    Messages:
    6,259
    Something not mentioned yet in this thread: machine learning.

    This is "intelligent" enough nowadays that face-recognition in places like airports etc, is fairly widely used. Oh yeah, in London, the city with the most surveillance cameras, software can track an individual by analysing how they walk.
    Machine learning and software that can rewrite itself are perhaps the scariest things I can mention about AI. Scary because of the level of trust the authorities have now, and will have in the future as the technology "matures".

    Ok. So would you get in a driverless taxi? A bus or a train with no human driver? How about an an aircraft with no pilot?
     
  2. Google AdSense Guest Advertisement



    to hide all adverts.
  3. parmalee peripatetic artisan Valued Senior Member

    Messages:
    2,596
    ... and a bunch of stuff about Skynet and T-9000s and whatnots.

    Whoops, my bad! Those were Bob-a-Builder's ridiculous claims. I've seen neither you, nor anyone else within this thread, expressing grave concerns over their programmable toasters or video game systems usurping their authority and autonomy.

    I've no doubt that he can find such nutters on the web, but he seems to be in the wrong place if he wishes to find anyone here expressing concerns along those lines.
     
  4. Google AdSense Guest Advertisement



    to hide all adverts.
  5. parmalee peripatetic artisan Valued Senior Member

    Messages:
    2,596
    I believe Billvon (or perhaps someone else) introduced NPUs and the like.

    Bob-a-builder seems to be struggling with the concept of intelligence itself, and how such may be defined within different contextual/conceptual frameworks. Sheesh, I've used a pedal based drum machine which has been touted as the first "intelligent drum machine"--Digitech's SDRUM <<<--and it's really not an unreasonable claim, as it's got the capacity to learn (or "learn") within a very limited context.
     
  6. Google AdSense Guest Advertisement



    to hide all adverts.
  7. Michael 345 New year. PRESENT is 69 years old Valued Senior Member

    Messages:
    7,710
    I am started to worry my toaster is trying to convert me to religion

    Three times last month it popped up toast with Jesus face and twice for the virgin Mary

    Please Register or Log in to view the hidden image!

     
    parmalee likes this.
  8. parmalee peripatetic artisan Valued Senior Member

    Messages:
    2,596
    I would only start to worry if you get an image of Paul!
     
  9. Michael 345 New year. PRESENT is 69 years old Valued Senior Member

    Messages:
    7,710
    Why would I worry about Paul McCartney?

    Please Register or Log in to view the hidden image!

     
  10. James R Just this guy, you know? Staff Member

    Messages:
    32,775
    Bob-a-builder:

    Wrong. For example, I know various computer languages and I don't think it's nonsense.

    Depends what you call thought. There's certainly computation going on - just like it does in your head. Calling one "thought" and the other not-thought is just a matter of labelling. Unless you have a rigorous definition of "thought", that is.

    My computer runs a whole heap of applications that are much more complicated that turning lights on or off. For instance, it runs the web browser that I'm typing this post on. It can talk to the internet. It can keep time. It can play complicated games. And lots more, besides.

    It's like you saying that your brain can only do one thing: fire neurons or not fire them. You would conclude, on that basis, that you're not intelligent, by the same argument you're trying to use here.

    What do you mean by practical?

    What's the fundamental difference between your "wetware" brain and, say, a neural network based in silicon? Why is your brain "hardly a computer"? Please explain.

    Nonsense. You mentioned chess-playing computers, for instance. They emulate thinking about how to win a chess game very well, as far as I can tell. In fact, I'd even go further that that. They don't just emulate thinking about how to win a chess game. They actually think about how to win a chess game. If that's not what's going on, tell me what a computer is doing to win a chess game. Obviously, whatever it is doing, it must be somehow evaluating whether one move is likely to be better than another. If that's not thinking, I don't know what is.

    The boxes often have more complicated instructions in them than that.

    Again, it comes back to what you want to call "thought". Maybe you ought to try to define what you mean by thought. Considering possible future moves and calculating both sound like kinds of thought to me.

    Is that how long you've spent?
     
  11. Bob-a-builder Registered Senior Member

    Messages:
    119


    No. Your computer does not just "run your server" in a single process. Even if you were to use a "one minute timer" inside a program you don't just tell the computer. Let me know when a minute is up although modern code emulates such. The computer adds or subtracts from a common variable until it reaches the timing variable equal to the length of one minute.

    In a single timer a computer may repeat the same comparison 1000 times (example). Until x>999. Adding one to the value at a time.


    I meant. Name one application for a self thinking computer. If you wanted a computer to operate your server and it decided instead to play virtual chess with itself. Your work would not get performed.

    If you wanted a computer to calculate 1 plus 1, and it decided to argue about philosophy instead. You would be better off counting to two on your fingers. An Abacus does not have free will.

    That is what I meant by no practical purpose. It might be interesting or even fun to have a computer that is capable of more than comparing things, but name one purpose in the society we have today. Sure it can design art the same way it can "play" chess. But if you wanted a computer to do art or "play" chess and it instead decided to discuss your love life. That is not practical.

    Tell me ONE (1) line of coding that a computer should independently ignore when following software. There is none. If you design a computer and software. Every line needs to be followed in sequence or it is "not practical" for any conceivable purpose.

    That is what I meant by "Would not be practical?". Even if the random "thinking" (seeming) were to arrive at thoughts similar to a human through algorithms. It would not serve any function we know of. You wouldn't want one in your GPS.

    Yes. You and others have already made it clear you perceive our minds as exactly the same as computers. If you have proof of this woo I would be happy to consider it. A journal link perhaps?

    The main argument I have seen thus far from Billvon and yourself is that thought is undefined. You seem to feel it is just comparing two variables over and over and then pointing to a switch (although you failed to grasp that they could point to subroutines which also run on comparisons and pointers to switches).

    I have not suggested it merely turns a lamp on and off as you seem to strawman. I have made that clear in many earlier postings. Every light on your screen however is the result of one such subroutine forming what we view as images.

    It would seem pretty daft to me that any company would even finance a self thinking computer that could not follow software. It would have no use except to tell us we are inferior if so.


    Again; that is comparing every possible answer and comparisons based upon points about which is the best move.

    A very powerful computer could project the outcome of every possible move going forward. A weaker program may calculate less moves ahead but choose a moved based upon points.

    Where we can say "Checkmate in 6 moves" based on experience and our own projections, a computer could calculate it will be victories many moves earlier.


    Yes. You and others have made it clear you consider thought to be nothing more than comparing 2 items at a time. Similar to what computers do. This is a unique argument and kudos (Billvon proposed it before you though), but I'm unsure I've seen anything resembling actual proof of that.

    However; I had conceded to Billvon that if he considers a brain as just neurons that can only judge a minute by running subroutines and adding neuron values similar to "if then else" statements. Then your definitions of thought are indeed similar to that of a computer. We could break "love" for example into variables and processes.

    "Every line of functional (added that word for Billvon) computer programming in EVERY COMPUTER LANGUAGE is nothing more than an "if then else" statement or turns a light switch (or further subroutine in most cases) on.

    Ad Hominem is not valid science. If you have any documented proof of this woo I would be happy to consider it or debate further.

    AI is woo based upon fearmongering from movies and computers beating chess masters by calculating every possible move or to best point based situations.

    Before this thread I had not considered some would consider our minds as limited to the exact same two processes. Diamond flowchart decision boxes, or square pointer boxes. It appears many here do. I think that is novel but would prefer actual science.
     
    Last edited: Jun 19, 2019
  12. billvon Valued Senior Member

    Messages:
    15,857
    You are conflicted, then.

    You admit that all neurons can do is either fire or not fire. Very similar to computers. Yet you claim that humans have intelligence, but computers cannot have intelligence because they work just like neurons. That is a contradiction.

    Science tells us that despite both being based on binary processes, both can exhibit intelligence.
     
  13. spidergoat Venued Serial Memberlist Valued Senior Member

    Messages:
    53,768
    No one would assert that present day computers have a general intelligence. But we have proved that they can develop their own algorithms to do specific things better than a human, and in a way that we don't completely understand. Neural networks mimic the way biological brains work, albeit with less complexity and versatility. This points to the idea that in the future, neural networks could teach themselves general intelligence, as well as the ability to perceive and manipulate the physical world. The difference between computers and biology is less a matter of essential differences but degree. This will become more evident when computers are built that are closer to brains in structure, with the innate ability to weigh more than one variable at a time. It's a theme of science fiction whether a sufficiently advanced artificial structure can actually be sentient and how would we know. Its an open question. Is there a difference and what is the difference other than materials? What is the test for consciousness?
     
  14. Bob-a-builder Registered Senior Member

    Messages:
    119
    No. Suggesting all brains do is fire neurons is your claim which I would consider if you provided actual research to back up your incredible claims.

    That is your woo. Not mine. I apparently have enough because I do not buy into History as dictated by the church.

    Every cell in our body is influenced by the foods we eat and emotions we express. Trillions of these neurons you speak of do exist but they adapt and change routes faster than we can calculate and you suggest it is all logic based. Perhaps that is true. It just has not been proven.

    I agree that YOU and a few others THINK that is how a brain operates. It is not my suggestion, so please do not credit me with that fallacious assumption. It may be true but I do not yet agree. I would consider your assumptions (not valid science) if there is actual science behind it.

    You are positing our brains are computers and that every emotion is binary based. Perhaps it is true but unsure how that could cover original thoughts. Computers can only compare values. How could that ever equate to original thinking? (Not really asking)

    That is a fun conspiracy theory. Perhaps fiction books have been written on it.
     
  15. spidergoat Venued Serial Memberlist Valued Senior Member

    Messages:
    53,768
    You are the one making the assertion that has yet to be proven, that human thinking is more than the sum of its parts. Why do you think so? So far you have given the answer in terms of foods, emotions, and fast connections. Can you elaborate? What's the difference between thinking and comparing values? Axons compare electrochemical values and yet the collective result is thought.
     
  16. billvon Valued Senior Member

    Messages:
    15,857
    Just like suggesting all computers can do is compare two values or turn something on or off.
    The basic operation of every single neuron in your brain is basically the same. Fire or not fire.

    Now, how often do they fire? How many of them fire at once? What threshold do they fire at? Where is that action potential transmitted to? All those things are affected by things like the food we eat, the input we get and what we learn.

    Again, just like computers. The basic operation of a conventional computer is to flip bits on or off. Now, which bits do they flip? How often do they do it? What changes the pattern those bits form? What is done with that information? All those things are affected by programming, the machine's inputs, its training and the patterns it sees.

    For example, I am working on a neural processor. At its most basic it's a bunch of bits that turn on and off. But load it up with training code and teach it to recognize faces and it will be able to find your old girlfriend's face in the thirty thousand pictures you have on your hard drive. Or it will be able to talk to you in plain English. Or it will be able to drive your car for you. Or beat you at chess. Or figure out when your wife is pregnant (without any medical input whatsoever.) Or figure out what disease you have, or where you're from, or what college your kids will go to.

    (BTW fun fact - NPU's aren't programmed. They are trained.)
    If "a few others" mean "scientists and doctors" then yes, agreed.
    Absolutely not. I am saying that the neuron, which is the basic building block of every cognitive activity you engage in, is on-off. Very much like a computer. But have enough of those on-off neurons, interconnected in uncountable ways, and you get Picasso, Einstein, Teller and Hawking.
    Neurons can only fire or not fire. They lead to original thinking. So far neural networks are about as complex as a fruit fly's brain, so don't expect to see much original thinking from them yet. But given that computational capability is still doubling every 18 months, that will improve very quickly.

    (from a Quora post:

    In most neurons, propagation of an electrical signal occurs via "all" or "none" action potentials. These neurons integrate their input stimulus and fire when the input stimulus exceeds a certain threshold. This form of information transmission, driven by voltage gated ion channels, ensures rapid and robust signal transmission over long distances. The output of these neurons(with chemical synapses) is a burst of neurotransmitter release driven by the action potential.)
     
  17. Baldeee Valued Senior Member

    Messages:
    1,702
    If computers could not come up with original actions/outputs, how are computers among the greatest ever chess and Go players, able to out-"think" other human players, come up with tactics that human players have never seen before?
     
  18. TheFrogger Valued Senior Member

    Messages:
    1,596
    I agree. True machine intelligence is impossible. But here's a thought: what if a human thought like a machine? Would that make the person robotic? Could a chess player consider every possible future move??

    Please Register or Log in to view the hidden image!

     
  19. spidergoat Venued Serial Memberlist Valued Senior Member

    Messages:
    53,768
    How do you know that artificial intelligence is impossible? What is "true" intelligence?
     
  20. TheFrogger Valued Senior Member

    Messages:
    1,596
  21. Michael 345 New year. PRESENT is 69 years old Valued Senior Member

    Messages:
    7,710
    Thought bubble

    Would Artificial Intelligence have enough Intelligence to know it was Artificial?

    Please Register or Log in to view the hidden image!

     
  22. spidergoat Venued Serial Memberlist Valued Senior Member

    Messages:
    53,768
    I don't see why not, it would be a public fact. Even if not, the evidence of manufacture would be evident, printed circuits don't occur in nature.
     
  23. TheFrogger Valued Senior Member

    Messages:
    1,596
    But sand does (silicon (silly-con) is MADE of sand.) Electricity was inspired by nature (lightning.)

    Please Register or Log in to view the hidden image!

     

Share This Page