Artificial Intelligence

[otokage007]

Wrong. Evolution is all about adaption due to different effects of an environmet. You could have the same species with different traits.

Well i've never claimed the opposite. So? :/

Lol the level of arrogance you posess is astounding considering how your own knowledge is incomplete and flawed. First of all evolution doesn't just 'invent' something. There are random mutations all the time. Depending on natural selection some mutations are picked over others. Now these mutations result in new proteins which are more adaptive. How these proteins create completely novel phenomenon extrinsic to the universe is impossible to explain. If you think they 'create' signals which invent these novel phenomenon then YOU do not understand evolution. Evolution does not just 'invent' mental phenomenon. And even IF it did, that still does not explain what these mental perceptions are! Trying to sound arrogant about it doesn't really help your case and actually shows how little you know about what I'm talking about.

Thanks for the class Dawkins lol. But the fact that you just don't know how to explain that "perception" existed since the very beginning of life and how by an evolutionary process this phenomenom evolved into the five senses we humans have, doesn't mean this is something metaphysical.

Even inert matter interact with the environment, it's called chemistry. Organics' perception is just this interaction after hundreds of millions of years of evolution. Nothing incredibly novel.

No.
Stop trying to show that you're logical because you're not. Even if bacteria evolved into preferring one biochemical aspect from another it still does not explain how the subjective experience of taste exists.

It explains that it is not a novel process, just an ancient process that aquired too much complexity for you to understand it.

I probably know more biology than you do

Despite the fact that I'm a biologist, well yes, you do. :sarcastic

stop trying to teach me things that I already know.

Well it was you who tried to teach me evolution, not vice versa.

But back to the topic. Everyone here is telling you that u haven’t shown any proof of the metaphysical things u claim. And you keep quoting yourself again, again, again and again, always showing the same sources that clearly don’t convince us. The only thing u have shown, is that taste, colors or sounds, are our brain’s invention and don’t exist out of our brain. Well, let me be clear so u finaly understand: WE ALREADY KNEW THAT. And that proves nothing. Now you are stuck because u see ilogical that evolution created such things, well, is that really a good reason to try to find a solution in the metaphysical?

 

[1137]

We will be able to create strong AI without the utilization of fairy dust.

Not with every single scientist in the world apparently becoming mystics...

If I'm right then yes. If I'm wrong and there's some biological code then no, strong AI is impossible. Unfortunately you don't realize that.

Lol, we get what you're saying, we're just saying and showing it is wrong. Recognize the difference.

 

[idav]

So the brain decodes light waves into light we see? Okay so you're saying that there's an algorithmic approach towards decoding the waves into an output? Okay..that still doesn't explain what the output is. I want you to explain through this 'code' how the brain creates perception.

If we knew we'd already be downloading our memories so we can save them to our hard drive. Once we crack it we will be able to express memories in computer terms as ones and zeros. There is an obvious language barrier.

And btw you are talking magic as well, because according to your theory, the brain evolved this code which started assigning variables to physical sensations. Where did the variables come from? The brain made it up? But our universe doesn't have such entities? Well that's the power of the brain? Lol.

The variables came from evolution in the form of DNA changes and mutations. How do you know the universe doesn't have such entities, it made us didn't it.

In computer terms red on the computer screen is a wavelength that is emitted by the computer which our brains translate as red. The computer does not create red. It conveys the information of a wavelength which we translate as red.

We evolved to perceive light same as plants. Light causes cells to give chemical and electrical responses that we interpret as a sensation. Same as any sense works, as a receiving and distribution of information.

 

[MD]

The variables came from evolution in the form of DNA changes and mutations. How do you know the universe doesn't have such entities, it made us didn't it.

Well this is exactly what I'm saying. The universe has such entities, but they are mental properties. The mental properties are the variables. If the universe has such mental entities then the universe has another a mental property to it, aside from matter and energy.

 

[LegionOnomaMoi]

Once we crack it we will be able to express memories in computer terms as ones and zeros.

Unless non-procedural memories (semantic, episodic, etc.) are the result of incomputable processes. As it seems increasingly plausible that a great deal of biology is not reducible a finite state machine, and that computable models of organisms or biological systems are approximations, the assumption that neural processes underlying the encoding, storage, and processing of "meaning" or concepts can be simulated via binary code is overly optimistic at best, and perhaps unwarranted.

 

[idav]

Well this is exactly what I'm saying. The universe has such entities, but they are mental properties. The mental properties are the variables. If the universe has such mental entities then the universe has another a mental property to it, aside from matter and energy.

How are humans something other than matter and energy? Why would biological lifeforms have some other substance that isn't found anywhere else in the universe except in life? The answer is that lifeforms are made of the same things as the rest of the universe and is a result of cause and effect since the beginning. Volition is just cause and effect with variables that aren't easily calculable but it doesn't mean it is impossible to calculate. With enough knowledge anything can be calculated.

 

[1137]

How are humans something other than matter and energy? Why would biological lifeforms have some other substance that isn't found anywhere else in the universe except in life? The answer is that lifeforms are made of the same things as the rest of the universe and is a result of cause and effect since the beginning. Volition is just cause and effect with variables that aren't easily calculable but it doesn't mean it is impossible to calculate. With enough knowledge anything can be calculated.

Perhaps our friend here doesn't understand how everything that exists can be caused by cause / effect since the big bang. I can say confidently that consciousness wasn't around when the big bang occurred, nothing that happened needs a mystical aspect to be explained. This idea that lacking full understanding = unexplainable is truly getting old.

 

[idav]

Unless non-procedural memories (semantic, episodic, etc.) are the result of incomputable processes. As it seems increasingly plausible that a great deal of biology is not reducible a finite state machine, and that computable models of organisms or biological systems are approximations, the assumption that neural processes underlying the encoding, storage, and processing of "meaning" or concepts can be simulated via binary code is overly optimistic at best, and perhaps unwarranted.

I've already seen that science is working on decoding of motor responses from neurons. There are also scientists working on decoding visual perception. The idea would be to have a sort of Rosetta Stone for computer language to brain language. If we are able to give or retrieve the appropriate firing sequences and the associated neurons for moving an arm then we should be able to do something similar for visual perception but obviously that would be much harder.

 

[LegionOnomaMoi]

I've already seen that science is working on decoding of motor responses from neurons.

Which is quite different from encoding or simulating.

There are also scientists working on decoding visual perception.

A study I'm currently involved in concerns the encoding and categorization of the shapes of visual stimuli.

The idea would be to have a sort of Rosetta Stone for computer language to brain language.

Which is not necessarily possible. For as long as computer science has existed, computer scientists have known of incomputable/noncomputable problems. What we are less sure of is how modeling or simulating real world phenomena may be impossible using modern computers (finite state machines which are equivalent to or "reducible" to Turing machines). In order for "a sort of Rosetta Stone" for brain language to computer language to exist, there must be a mapping or function from the "language" used by the brain onto that used by computers which does not result in so many elements y of the codomain produced by some common x of the domain that the resulting map was useless.

We can currently build models of neurons and models of neural networks. We can create programs which process language. However, the amount of approximation involved in models of neural activity makes these at best useful for understanding the brain, not simulating it. Nor is there any scientific reason to believe that the activity of the brain can, even in principle, be simulated by classical computers. Rather, the widespread belief that this is true (a belief which is still widely held, but now also widely doubted) resulted from a faith in the power of new and wondrous computational devices, a lack of appreciation of complexity and dynamical systems as well as the limits of our mathematical tools for dealing with nonlinearity, and finally the fact that we took to many tasks humans and some animals do so easily for granted. 50+ years later, a lot has changed.

If we are able to give or retrieve the appropriate firing sequences and the associated neurons for moving an arm then we should be able to do something similar for visual perception but obviously that would be much harder.

There is no "appropriate firing sequence" for moving an arm. Not just because the neuroscience community has yet to come to an agreement over comparatively "basic" neurocomputational issues (rate vs. temporal encoding), but also because there is no single group of neurons or groups of neurons which consistently fire in the same way to produce the same arm motions. The brain is far more dynamic.

 

[idav]

Perhaps our friend here doesn't understand how everything that exists can be caused by cause / effect since the big bang. I can say confidently that consciousness wasn't around when the big bang occurred, nothing that happened needs a mystical aspect to be explained. This idea that lacking full understanding = unexplainable is truly getting old.

I can understand the issue people have and goes into the issue you have with having awareness as a fundamental component. We can't really explain how we go from cause and effect to volition, the idea would be that cause and effect in itself is the first step to sensing which is what I take to mean inherent awareness.

 

[idav]

A study I'm currently involved in concerns the encoding and categorization of the shapes of visual stimuli.

Way cool!

There is no "appropriate firing sequence" for moving an arm. Not just because the neuroscience community has yet to come to an agreement over comparatively "basic" neurocomputational issues (rate vs. temporal encoding), but also because there is no single group of neurons or groups of neurons which consistently fire in the same way to produce the same arm motions. The brain is far more dynamic.

Never thought it would be easy but I think it's possible. The redundancy of the trillions of neurons does make for a difficult task. There is still logic to the brain even if is non-linear. A certain sequence from the neurons would mean move arm left or a picture of mom. Someone must think it's possible otherwise there would be no funding for scientists to try and figure it out.

 

[LegionOnomaMoi]

Way cool!

Only if it results in anything useable. Also, the novelty and "cool" factor of seeing scans of someone's functioning brain via an fMRI during some experiment wears off pretty quickly. The IRB doesn't even allow researchers to have fun by telling subjects "omg! there's something moving around in your skull! It's...it's...an alien life form!".

There is still logic to the brain even if is non-linear.

Depends what you mean by "logic".

A certain sequence from the neurons would mean move arm left or a picture of mom. Someone must think it's possible otherwise there would be no funding for scientists to try and figure it out.

They don't. Neuropsychologists tend to be interested regions of the brain and things like "substrates" and "neural correlates", while neuroscientists (particularly computational) may be interested in modeling individual neurons or neural networks or both, but not how particular neurons/networks are active in any particular arm motion. The more one works with how human cognitive systems work, the less one is dealing with actual neuroimagining and neural modeling. In particular, modeling individual neural properties is extremely difficult because most of the time things like intracellular measurements of a neurons activity are impossible. Additionally, most of the time (perhaps always) the spike train of an individual neuron has no independent meaning, as it conveys information only through e.g. its correlation with the spike trains of other neurons. So those of us who look at brain function and cogition along with biology tend to ignore individual neurons and often even neural networks, and instead focus on particular regions of interest (ROIs) and how they may be involved in a particular type of cognitive function. Computational neuroscience, on the other hand, is both much more focused on neural properties and much less about what neurons/neural regions are invovled in anything.

Nobody is trying to figure out how particular neural sequences correspond to particular actions/thoughts/memories/etc. Good thing too, as there isn't any consistent one-to-one correspondence.

 

[LegionOnomaMoi]

From Filo & Lotan's Information Processsing by Biochemical Systems: Neural Network-Type Configurations (Wiley, 2010), sect. 1.1 (emphases added):

"Both living organisms and computers are “information-processing machines” that operate on the basis of internally stored programs, but the differences between these systems are also quite large. In the case of living organisms, self-assembly occurs following an internal program, and the nervous system and brain formed in this way function as an autonomous information machine. Unlike traditional computers which must be “driven” from the outside, biological systems have somehow incorporated within them rules on how to function. Moreover, in the case of biological entities for which there is no external blueprint, the design plan is entirely internal and is thought to undergo changes both in the evolution of species and in the development of individuals. These similarities and differences have drawn the attention of computer scientists as well as of life scientists.

In order to revolutionize the current world of computers, three roads, or any combinations of them, are clearly visible

1. Changing the physical elements at the foundations of the computer components
2. Changing the architecture of computers
3. Devising new software and computing algorithms
It is, however, true that a biological computer (or biocomputer) of a completely different nature from today’s electronic computers already exists in the form of the fundamental phenomenon of life. The most advanced machinery, a living organism, operates with functional elements that are of molecular dimensions and actually exploits the quantum-size effects of its components. Yet the quintessentially biological functions of living forms: autonomy, self-organization, self-replication, and development, as witnessed in both evolution and individual ontogeny, are completely absent from current computing machines."

The authors of the above study attempt to defend the capability of artificial biocomputers, particularly in areas where classical computing has failed so spectacularly (e.g., simulating biological, including neural, systems). Similar works exist for Quantum computers. Although such works inevitably depend on a great deal of theoretical approaches and devices, along with a fair share of assumptions, at the very least the typically represent improvements to the assumptions underlying classical cognitive science and artificial intelligence research: hardware is irrelevant and everything is algorithms. This assumption came from the formalism of the 19th and early 20th century and the work of Turing, Church, and others on how formal systems could be implemented on "Turing machines" or anything equivalent to them. There was nothing behind the assumption that this formalism could be used to simulate any and all systems or that it was used by any and all systems in some sense. As our understanding of biological complexity and systems increases, along with our understanding of the neurophysiological properties of cognition, it seems increasingly inappropriate to apply this classical computer & cognitive science model to the brain (among other systems).

 

[idav]

As our understanding of biological complexity and systems increases, along with our understanding of the neurophysiological properties of cognition, it seems increasingly inappropriate to apply this classical computer & cognitive science model to the brain (among other systems).

All that is some good food for thought. As you mentioned quantum computing and as we go more into the realm of synthesizing biological functions I think it is perfectly appropriate. Our advances in computing and nanotechnology will only keep leading us closer to being able to replicate some of those things mentioned like autonomy, self organization, learning etc.

They say that by 2025 a $1000 computer will have the power of the human brain. This is based on the power of current supercomputers. Now we are just talking raw power not intelligence but it certainly would be a step in the right direction. They say you have to be able to run 10 trillions of processes in a second.

Computing power to equal human brain by 2025 | GulfNews.com

 

[LegionOnomaMoi]

They say that by 2025 a $1000 computer will have the power of the human brain.

One day I will find out the names behind this "they", track them down, and slap them all in the face.

Keep in mind that "they" also said we would have computers with the power of the human brain decades ago. The problem with what "they" say is that "they" keep being wrong. Maybe "they" just have bad P.R.

 

[idav]

One day I will find out the names behind this "they", track them down, and slap them all in the face.

Keep in mind that "they" also said we would have computers with the power of the human brain decades ago. The problem with what "they" say is that "they" keep being wrong. Maybe "they" just have bad P.R.

They are IBM. Mind you this article is from 4 years ago and the amount of increased power potential just in the four years is unbelievable. I will be very interested to see where we are at in 10 years, the progress is increasing exponentially as far as raw power and memory.

 

[Reptillian]

One day I will find out the names behind this "they", track them down, and slap them all in the face.

Keep in mind that "they" also said we would have computers with the power of the human brain decades ago. The problem with what "they" say is that "they" keep being wrong. Maybe "they" just have bad P.R.

Lol, I always tell my friends that I'm the "they" in "that's what they say". :run:

 

[LegionOnomaMoi]

They are IBM.

In this case, but I constantly hear (and am sometimes myself guilty of similar ascriptions) what some nebulous "they" has said about any given topic.

Mind you this article is from 4 years ago and the amount of increased power potential just in the four years is unbelievable.

And yet we are no closer to computational devices which capable of semantic, rather than syntactic, processing than we were decades ago. We've just gotten better at faking it.

I will be very interested to see where we are at in 10 years, the progress is increasing exponentially as far as raw power and memory.

Assuming that you mean exponentially in the non-technical sense, that's true. But most who work within computer science believe we're quickly approaching a limit.

 

[idav]

Assuming that you mean exponentially in the non-technical sense, that's true. But most who work within computer science believe we're quickly approaching a limit.

We always think that and we still manage to surprise ourselves.

 

[LegionOnomaMoi]

Lol, I always tell my friends that I'm the "they" in "that's what they say". :run:

Very interesting. On a completely unrelated note, if you would kindly supply (feel free to PM) your home address, schedule including maps and building schematics, and a list of your worst fears, that'd be great. Nothing to do with my intent on revenge against those behind what "they say", just a completely innocent request.