Map Makers Admit Mistake in Showing Ice Cap Loss in Greenland

[Matthew N. Wright]

How can climate change be controlled by science?

You don't watch enough cartoons. Don't you know that before Bitcoin, when people asked "how" all you had to say was "Science!". It replaced its long lived predecessor "Jesus".

[1713504163]

1713504163

[1713504163]

1713504163

[bb113]

Ok, why not go off topic... matthew thinks this thread is dumb anyway. In simulating neural activity (which I know little about), how detailed are the models of dendritic arbor and spines?

Warning, jargon below:
I mean do they attempt to model all the postsynaptic feedback factors (AMPA and NMDA receptors, various GPCRs and G-proteins, etc) to level of the cytoskeleton, or is it just calcium influx -> greater weight? Also, how is the weight of each synapse adjusted according to distance and number of branch points from the soma, is this corrected for dendritic thickness?

What would the effect on firing rate be if a certain treatment increased the proportion of distal dendritic branches while decreasing the number proximal to the soma (i.e. results of sholl analysis)? Assume same input firing rate, homogenous spine density, etc.

If you're genuinely interested, read everything you can about the Bluebrain project: http://bluebrain.epfl.ch/

They are using an IBM supercomputer which has enough resources to dedicate the equivalent of one laptop per neuron, each having perhaps 5,000 synaptic connections for the simulation of one cortical column. Watch the simulation of a cortical column completely built from scans of slices of a rat's brain: http://www.youtube.com/watch?v=xHi9oLzvD8E

Google PDF documents where Henry Markram is a coauthor. Watch the TED video he gave. Their goal is build a complete human brain. As for some of your technical questions, I don't have the answers.

I personally was implementing STDP and an empirical model to simulate a neuron's action potential, with very efficient storage of synaptic connections, simulating the length of axons and dendrites by storing them in a list, sorted by length, such that the program could traverse the list, adding in the delay from one to the next, and thus the program would always know the next synapse which would fire, and these pulses would accumulate to each respective receiving neuron, such that it could be calculated which neuron would fire next, thus distributing pulses further on down the line.

Ah, I actually have never heard about this project (or at least didn't pay attention if I did) but have consistently come across great papers out of Henry Markram's lab.

[bb113]

While on the subject of brains, from a philosophical viewpoint, do you buy Dennett's arguments, or do you prefer the ideas articulated by Chalmers? Or, do you think Hameroff's studies, inspired by Penrose are the ticket? While I think Hameroff is onto something, given his credentials and line of study, I think he only seems to be addressing the 'how', as opposed to the 'why'. It's Dennett's and Chalmers' sparring that really goes after the Hard Problem. As for me personally, while I admit that Dennett is a great writer, and his stuff is really interesting to read, in the end, his answers are just unsatisfactory, although he seems so insistent that he's answered the big question. That's why, ultimately, I find Chalmers' discussion on the subject to be the most interesting.

Honestly I stopped paying close attention to the "what is consciousness argument" a couple years ago (I think that is what you are referring to)... Real answers will come from generating data, philosophizing is just a fancy word for bullshitting, psychology is only slightly better... not that its useless. My position is that I am just as capable of coming up with logical sounding theories as anyone else until we have more data. I think the "political impediments" to studying hallucinogens is really hurting this type of research. I got a chance to have dinner with David Nichol's recently and it was awesome. His son has actually published some interesting stuff on free will as well, beyond that I haven't been paying attention.

[bb113]

How can climate change be controlled by science?

You don't watch enough cartoons. Don't you know that before Bitcoin, when people asked "how" all you had to say was "Science!". It replaced its long lived predecessor "Jesus".

Agreed, there is a disturbing case of science=fact in some circles. It misses the whole point.

[FirstAscent]

While on the subject of brains, from a philosophical viewpoint, do you buy Dennett's arguments, or do you prefer the ideas articulated by Chalmers? Or, do you think Hameroff's studies, inspired by Penrose are the ticket? While I think Hameroff is onto something, given his credentials and line of study, I think he only seems to be addressing the 'how', as opposed to the 'why'. It's Dennett's and Chalmers' sparring that really goes after the Hard Problem. As for me personally, while I admit that Dennett is a great writer, and his stuff is really interesting to read, in the end, his answers are just unsatisfactory, although he seems so insistent that he's answered the big question. That's why, ultimately, I find Chalmers' discussion on the subject to be the most interesting.

Honestly I stopped paying close attention to the "what is consciousness argument" a couple years ago (I think that is what you are referring to)... Real answers will come from generating data, philosophizing is just a fancy word for bullshitting, psychology is only slightly better... not that its useless. My position is that I am just as capable of coming up with logical sounding theories as anyone else until we have more data. I think the "political impediments" to studying hallucinogens is really hurting this type of research. I got a chance to have dinner with David Nichol's recently and it was awesome. His son has actually published some interesting stuff on free will as well, beyond that I haven't been paying attention.

I think you're failing to appreciate the Hard Problem, and discussion of it. A lot of people confuse qualia with stimulation.

[bb113]

Probably, need more data before I'm interested.

[FirstAscent]

Probably, need more data before I'm interested.

You mean you're just not interested - which is fine. Your requirement for more data is a bit silly though. Perhaps you haven't read Chalmers and Dennett?

[bb113]

Dennet like 5 years ago, not sure about the others you mention. Don't get me wrong I love that stuff, but feel it is not really the most productive use of time. I went the path of philosophy -> pyschology -> neuroscience -> molecular biology (pharmacology). Basically I became dissatisfied with a top-down approach to understanding the brain/consciousness. It is too easy to generate logical narratives if you are unconstrained by data, then people start arguing about minutia and ambiguous definitions. But maybe I should get into this big picture perspective again. Here is my own concept (not sure how it matches up with what is in the literature):

The brain: An organ that receives information about its surroundings and outputs behavior. It has evolved into a way for large, long-lived organisms to generate a wide set of novel behavior responses to novel (or at least an uncertain set of) environmental conditions. Its value is that it allows predicting the future based on past experience. Although I use the term predictive, the brain's role here is as reactive decision maker. For organisms filling the large, long-lived niche, it is more energy efficient (from the perspective of the of genetic material) for behavior to be informed by past experience than the alternative: throwing a lot of hardwired variations on behavior at every problem and letting the best solution get selected out over the course of generations. This latter solution is predictive, but less reliable in the face of uncertainty.

The self: A subset of the environment that the brain receives a constant stream of information about (Do you consider your hair a part of yourself?) and is by far the easiest thing for the brain to predict the behavior of. The nature of the information from the self is also qualitatively different from that received about other parts of the environment.

Qualia: The simultaneous processing of multiple memories + sensory input. You will no more be able to completely describe a qualia with language than you can a picture. The parallel basis of qualia preclude description with a serial flow of information like language. It may be possible to communicate qualia via other methods though.

Consciousness: An awareness of the current and (projected) future self. I wouldn't say people are "conscious" when panicked, for example.

Free will: Perhaps it can be described as an illusion. Before making a decision (for example what to buy at the pop machine), all information about the environment is incorporated up until milliseconds before the behaviour occurs, so imagining your response to this scenario beforehand can never be done using complete information... It is how the brain understands that its own past behaviour will not be a perfect predictor of future behavior. The ability to incorporate this reality into "consciousness" is necessary to facilitate complex social interactions. Attributing free will to a "wrong-doer" makes it easier to ignore feelings of empathy by underestimating the role of circumstance in the wrong-doers behavior. Turning this back on yourself is the brains way of explaining behavior modification in response to negative social signals.

Critique away  

[FirstAscent]

You clearly align yourself with Dennett. Dennett attempts to explain the existence of qualia by explaining the resultant effects of physical processes happening inside the brain. It's not a satisfying explanation. I prefer Chalmers.

The first step: be very clear on what qualia is. It isn't the perception of red. It isn't even the assigning of a label of red in your mind, i.e. "I see the color red", which is in fact the recognition of seeing red and consolidating of all that neural activity into an alternate, perhaps simpler set of neural activity: symbolism via neural circuitry. Qualia are a set of experiences which equate to what it's like to see red.

Consider: the brain is a parallel machine. It is composed of cells which fire together through internal and external feedback. However, the resultant effects of those processes can be simulated on a computer either via parallel methods with the appropriate hardware, or serial hardware.

In other words, given a machine that has one processor and enough memory, the activities of neural circuitry within a brain can be simulated - each and every neuron and its attendant dendrites and axons and action potentials. The program code can be tuned through equations which simulate physical processes to produce the exact same neural activity, and if the machine is fast enough, at the exact same speed.

The problem is, from a complexity view, the silicon machine can have the exact same architecture as a computer from ten years ago or earlier. Any Turing complete machine, serial or otherwise, given enough memory, can reproduce both the output of a brain, and the internal states of a brain. If you were to argue that that machine, while running the above described simulation, was experiencing qualia, then you must consider that that machine, while running any program, including its native operating system, is experiencing qualia. But it's perfectly reasonable and probably correct that a machine running serial machine code is not experiencing qualia.

Then we have panpsychism. Let's assume we live on a very large planet with many nations like China. Let's assume the population is equivalent to the number of neurons inside the brain. Either because that's the way it happens, or because the people are instructed to behave a certain way, let's assume that through communication among people, messaging occurs between the people such that messages are passed between people in patterns that exactly replicate what happens in your brain when you see red.

Ignoring the consciousness of each individual person in the above scenario (in fact, we can assume they are simple mechanical machines), can we say that the population as a whole is experiencing qualia?

This then naturally raises the question: is our world economy conscious? It certainly mimics the processes of the brain through the actions of its constituent elements and outputs behavior.

Yes, in a sense, we do need more data, but what will it get us? Likely nothing. Understanding the neuron down to the quantum level only allows us to create simulations at a more granular level.

What is needed is a theory on consciousness that is as revolutionary as Einstein's Theory of Relativity, which can make predictions, and allow for testing.

However, testing the behavior of the system which is allegedly conscious by observing what it does and says is not the correct way, unless you accept panpsychism - i.e. consciousness in everything, which may or may not be true. A new theory most likely must posit that consciousness is a fundamental property of physical matter (panpsychism in a sense), or an external as of yet unknown component of the Universe which exists, but we just don't know it yet.

[bb113]

The first step: be very clear on what qualia is. It isn't the perception of red. It isn't even the assigning of a label of red in your mind, i.e. "I see the color red", which is in fact the recognition of seeing red and consolidating of all that neural activity into an alternate, perhaps simpler set of neural activity: symbolism via neural circuitry. Qualia are a set of experiences which equate to what it's like to see red.

Yes, in a sense, we do need more data, but what will it get us? Likely nothing. Understanding the neuron down to the quantum level only allows us to create simulations at a more granular level.

Anything occurring at the quantum level is probably less important than understanding the structure and dynamics of the circuitry

In other words, given a machine that has one processor and enough memory, the activities of neural circuitry within a brain can be simulated - each and every neuron and its attendant dendrites and axons and action potentials. The program code can be tuned through equations which simulate physical processes to produce the exact same neural activity, and if the machine is fast enough, at the exact same speed.

What makes you say this? I don't think this is true at all.

[FirstAscent]

The first step: be very clear on what qualia is. It isn't the perception of red. It isn't even the assigning of a label of red in your mind, i.e. "I see the color red", which is in fact the recognition of seeing red and consolidating of all that neural activity into an alternate, perhaps simpler set of neural activity: symbolism via neural circuitry. Qualia are a set of experiences which equate to what it's like to see red.

Yes, in a sense, we do need more data, but what will it get us? Likely nothing. Understanding the neuron down to the quantum level only allows us to create simulations at a more granular level.

Anything occurring at the quantum level is probably less important than understanding the structure and dynamics of the circuitry

In other words, given a machine that has one processor and enough memory, the activities of neural circuitry within a brain can be simulated - each and every neuron and its attendant dendrites and axons and action potentials. The program code can be tuned through equations which simulate physical processes to produce the exact same neural activity, and if the machine is fast enough, at the exact same speed.

What makes you say this? I don't think this is true at all.

Why have you boldfaced "symbolism via neural circuitry"?

As for your two remarks, oddly they are contradictory.

[bb113]

I boldfaced to emphasize the importance of circuitry rather than quantum effects. I don't think they are contradictory statements, I just do not understand:

1) Why serial processing with large enough memory can be a proxy for parallel processing (I may not understand the role of the large enough memory here, so please explain)

2) How such a system would be able to replicate the dynamics.

Neural circuitry is not at all static. See this Video of spines/filopodia changing over the course of hours. So, any simulation would also have to account for these dynamics, and alter the algorithms with respect to previous activity as time passes. That is in addition to all the receptor desensitization and trafficking, etc. If you take the simulation out long enough it should also account for the susceptibility of certain cell types to different types of damage that occurs, e.g. mutations (due to differing epigenetic patterns, oxidative stress), etc as the brain ages. Astrocytes, microglia, and the vasculature also play crucial roles in maintaining and altering neural circuitry.

So, I have my doubts that any static hardware (even using accurate dynamic algorithms) could be used to properly mimic a system like this. I am not saying these models are not useful, just that a "silicon brain", as you put it, would be a fundamentally different phenomenon. Therefore, any experience this silicon brain had of qualia would necessarily be fundamentally different.

Essentially, the term qualia refers to an emergent property of a certain type of complex system (the brain). It is the way the brain responds to sensory input. If we expand this to include silicon brains and countries, we are saying that qualia refers to an emergent property of a complex system that arises due to the system "sensing" a change in it's environment.

According to that definition, any suitably complex system that changes in response to it's environment will experience qualia, and can be deemed "conscious". Suitably complex can reasonably be defined as "equal to or more complex" than the human brain. So, then consciousness is a word we use to describe the complexity and responsiveness of a system to it's environment; It is not a phenomenon, but a poorly defined measure of degrees of freedom. Consciousness is a definition, definitions are inherently tautologies.

[bb113]

Hmm, so is any suitably complex system "self-aware"? I would say yes. Self-aware in the same way as a human? No. Are all humans equally "self-aware"? I would say no, unless we define humans as self aware, and self awareness as a human trait. Then we are back to the tautologies.

Anyway, my point is that only human brains can be conscious in the way human brains are and so cannot truly be replicated in silico, if we expand the definition to include other complex systems, we are simply talking about degrees of freedom.

[bb113]

And oh yea. So then the real question people are searching for the answer for is "What are the components of the human brain and how do these components interact to give rise to the observed degrees of freedom?" The answer is "we don't know, need more data". If you would like a better answer do neuroscience research or donate the proceeds of your other efforts to the study of neuroscience. I suppose the philosophers are contributing to the effort by attempting to qualitatively describe the degrees of freedom, but a real answer requires quantification, which requires more data.

[FirstAscent]

I boldfaced to emphasize the importance of circuitry rather than quantum effects. I don't think they are contradictory statements, I just do not understand:

1) Why serial processing with large enough memory can be a proxy for parallel processing (I may not understand the role of the large enough memory here, so please explain)

How much do you know about single taped Turing machines? Nothing could be more serial. And they are capable of replicating the behavior of any computer, parallel or serial.

2) How such a system would be able to replicate the dynamics.

Neural circuitry is not at all static. See this Video of spines/filopodia changing over the course of hours. So, any simulation would also have to account for these dynamics, and alter the algorithms with respect to previous activity as time passes. That is in addition to all the receptor desensitization and trafficking, etc. If you take the simulation out long enough it should also account for the susceptibility of certain cell types to different types of damage that occurs, e.g. mutations (due to differing epigenetic patterns, oxidative stress), etc as the brain ages. Astrocytes, microglia, and the vasculature also play crucial roles in maintaining and altering neural circuitry.

Any Turing machine (serial) could replicate all of the above, to any degree desired. Memory is the only limitation.

So, I have my doubts that any static hardware (even using accurate dynamic algorithms) could be used to properly mimic a system like this. I am not saying these models are not useful, just that a "silicon brain", as you put it, would be a fundamentally different phenomenon. Therefore, any experience this silicon brain had of qualia would necessarily be fundamentally different.

It's difficult to answer your statement above when it is in conflict with what I have been saying, but then, in the last sentence, it concludes what I concluded. Your above statement, in the first sentence, is not well grounded. See my above statements about Turing machines. Your conclusions, however ill founded they are, basically say what I said. That is, one doubts the sense of such a machine being conscious. Unfortunately, your statements below confound the issue due to your incorrect assumptions as to what consciousness is and what qualia are.

Essentially, the term qualia refers to an emergent property of a certain type of complex system (the brain). It is the way the brain responds to sensory input.

You are completely wrong here. That is not what qualia refers to. The way the brain responds to sensory input is not qualia. What you have described is a physical process. Qualia refers to the accompanied experience of the way things seem which correlates to different processes occurring within the brain - typically a distributed activation of a subset of the brain's neurons. You might want to get familiar with the concept of Philosophical Zombies and their conceivability. Whether you buy into the conceivability of Philosophical Zombies or not is irrelevant - you still must understand what they are to fully discuss qualia. Note that we're not even talking about the possibility of Philosophical Zombies actually existing, merely the conceivability of them.

If we expand this to include silicon brains and countries, we are saying that qualia refers to an emergent property of a complex system that arises due to the system "sensing" a change in it's environment.

Again, not quite right, but closer. You must detach qualia from the term "sensing". Sensing refers to a change in state due to incoming stimulation. That is not qualia. Qualia is the the experience which accompanies sensing either internal or external changes in state.

According to that definition, any suitably complex system that changes in response to it's environment will experience qualia, and can be deemed "conscious".

Nobody is really making any such claim. If it were so simple, we wouldn't have such discussions, and people such as Chalmers, Dennett, Penrose and Hofstadter would be writing basic texts of how things work, rather than how they hypothesize what might be happening.

Suitably complex can reasonably be defined as "equal to or more complex" than the human brain. So, then consciousness is a word we use to describe the complexity and responsiveness of a system to it's environment;

Wrong. Consciousness is not a word we use to describe the complexity and responsiveness of a system to its environment.

[bb113]

Any Turing machine (serial) could replicate all of the above, to any degree desired. Memory is the only limitation.

Sorry for the multiple posts... Anyway, please explain this further.

Qualia refers to the accompanied experience of the way things seem which correlates to different processes occurring within the brain - typically a distributed activation of a subset of the brain's neurons.

Also, are you claiming that experience is not a physical process? If so, then this is the ambiguous definition we now need to argue about.

[FirstAscent]

Any Turing machine (serial) could replicate all of the above, to any degree desired. Memory is the only limitation.

Sorry for the multiple posts... Anyway, please explain this further.

Turing machines (and the concept of Turing completeness) are fundamental concepts within the fields of computer science and mathematics: http://en.wikipedia.org/wiki/Turing_completeness

Qualia refers to the accompanied experience of the way things seem which correlates to different processes occurring within the brain - typically a distributed activation of a subset of the brain's neurons.

Also, are you claiming that experience is not a physical process? If so, then this is the ambiguous definition we now need to argue about.

Physical processes are just that: tires rolling down the road, lightning strikes, soil erosion, breathing, photons hitting cells in you eyes, neurons firing. It is not appropriate to discuss consciousness within the context of those processes. Accept them for what they are: physical processes.

Qualia is the experience you seem to have inside your mind - the constant streaming movie and awareness of your life that you are witness to.

Do not confuse the experience of seeing red with the physical process of your brain processing the event precipitated by photons of the red wavelength hitting your eyes. Qualia is the experience which accompanies the physical process inside your brain. It is not the physical processes that are happening within your brain. I cannot stress enough the importance of becoming familiar with the concept of Philosophical Zombies at this point.

[bb113]

I am somewhat familiar with philosophical zombies and turing machines. I just learn best by asking questions... So, with regards to experience: Would you say "experience is the result of physical processes"?

[FirstAscent]

I am somewhat familiar with philosophical zombies and turing machines. I just learn best by asking questions... So, with regards to experience: Would you say "experience is the result of physical processes"?

That is the 64 million dollar question. The short answer is: yes.

The long answer:

1. Can we say that experience is strictly the result of physical processes, or is it the result of physical processes and some other component/layer/plane of the Universe that we don't know about?

2. Can we say that all physical processes give rise to experience, or only some types. i.e. is calcium a required component?

3. Is experience not tied to physical processes at all, but the timing of information transmission and in certain amounts at certain frequencies?

More information about what is happening physically inside the brain is unlikely to yield satisfying answers. More information will tell us how better to correlate physical brain processes to experience (i.e. consciousness, qualia and experience), but it will not answer the big question without a revolutionary theory, in my opinion.

Is consciousness a fundamental property of the Universe - something which exists and is ready to manifest when the proper physical structure and events occur? If so, what is that fundamental property?

[bb113]

I am somewhat familiar with philosophical zombies and turing machines. I just learn best by asking questions... So, with regards to experience: Would you say "experience is the result of physical processes"?

That is the 64 million dollar question. The short answer is: yes.

The long answer:

1. Can we say that experience is strictly the result of physical processes, or is it the result of physical processes and some other component/layer/plane of the Universe that we don't know about?

2. Can we say that all physical processes give rise to experience, or only some types. i.e. is calcium a required component?

3. Is experience not tied to physical processes at all, but the timing of information transmission and in certain amounts at certain frequencies?

More information about what is happening physically inside the brain is unlikely to yield satisfying answers. More information will tell us how better to correlate physical brain processes to experience (i.e. consciousness, qualia and experience), but it will not answer the big question without a revolutionary theory, in my opinion.

Is consciousness a fundamental property of the Universe - something which exists and is ready to manifest when the proper physical structure and events occur? If so, what is that fundamental property?

To rephrase in scientist language (correct me if I have misinterpreted):

1. Are known physical processes sufficient to explain experience?
- To answer this we must add all possible combinations of physical processes to an unconscious substance until it becomes conscious or we run out of processes to add.

2. Are all physical processes known to be involved in generating experience actually necessary to generate experience? Are there other processes that could substitute?
-To answer this we need to remove physical processes from a conscious substance one by one (and in all possible combinations) then record whether or not consciousness remains. Next we need to substitute all other physical processes (one by one and in combination) for each of the removed processes (one by one) and observe if consciousness reappears.

3. The timing and magnitudes of information transmission are not tied to physical processes. Do these phenomena explain consciousness?
-You are making the assumption that the timing and magnitudes of information transmission are not tied to physical processes. We need to address why this assumption was made.

It is impossible to prove a negative, one can only state a theory is exceedingly implausible when compared to alternatives. Is there any evidence for some other plane of the universe that contributes to consciousness? How strong is this evidence relative to the evidence that experience is strictly the result of physical processes? Since we will never have complete information about the universe, in the end there will always be an aspect of subjectivity (i.e., more spiritual people will give the unknown process hypothesis a higher prior probability). All we can do is look for support of the hypothesis that experience can be accounted for due to known physical processes. In other words, need more data.

The experiments mentioned above would get us a conclusive answer but they are infeasible to perform in a human lifetime (likely taking millenia), and we humans want answers immediately. So, rather than swapping in and out physical processes in a random order, we should learn more about what processes are important in generating experience to determine what the most likely physical processes may be. Then do the experiments with those likely culprits first. This is what neuroscientists are currently doing. Maybe humans are one of the experiments of an extremely long lived entity trying to rule out non-physical explanations for consciousness?

With regards to #3:
What examples of information transmission are you aware of that occur without being tied to physical processes?