Main Page

From O'Reilly
Jump to: navigation, search

Randall C. O'Reilly's Wiki

  • See Random Stuff for non-philosophy random stuff (software issues, gadgets, rc helicopters, misc sharing of tips etc)
  • See New Issues for a chronology of new additions and issues (blog-style).
  • See Extreme Moderate for my personal political philosophy.

This wiki helps me organize my personal thoughts about life, the brain, the universe, and everything. It is a 40th birthday present to myself, to take the time to get this stuff down on bits. The content includes all the topics you're not really supposed to talk about, including politics, religion, death, etc. Probably the reason you're not supposed to talk about this stuff is that inevitably various people will be offended. I apologize profusely in advance for all the strongly-worded offensive things in here, and extend my sincerest desire that we all come together as one big in-group. I have tried to be as constructive and inclusive as possible, while still remaining true to my beliefs. I would encourage those who take offense to engage in constructive dialog, or simply look the other way. My desire to share overwhelms my fear of offending! As this is a wiki, I encourage people to add stuff where relevant (especially pointers to other relevant ideas), but I reserve the right to edit and remove inappropriate material. Everything written by me is Copyright 2007, Randall C. O'Reilly.

Central Nodes in My Epistemological Network

It is tempting to try to develop a world view that is provably correct, following logically from a small set of axioms, or something nice like that. I believe this has been tried many times (e.g., logical positivism), without success. My own view of epistemology (how knowledge is structured and comes to be) is based on what we know about the brain: it is a neural network. Thus, our knowledge is organized like a network: with many interconnected nodes or beliefs. Maybe there is some kind of rough hierarchical structure, but it is not precise. This network is like a geodesic dome -- its strength depends on all the interconnections among the nodes. It has no clear starting point, no set of axioms from which everything else logically follows. I cannot tell you exactly how it came to be, or convince you that it is inevitably correct: I can only try to reproduce it here on this set of interconnected wiki pages, and see how well it resonates with your experiences.

As a network, the most important quality of these nodes is that they are self-consistent. Each belief or idea is compatible with all of the others, and with the total of my experiences in this life. Maximizing consistency (i.e., constraint satisfaction) is the best we can hope for in shaping our beliefs. I have found the following set of beliefs to be maximally self-consistent, and consistent with well-accepted scientific facts. Thus, to the extent possible, these beliefs should be true.

I Think Therefore I Am

Descartes got something right: Cogito Ergo Sum -- I think, therefore I am. The one thing I know for certain is that I am here thinking. I might be a brain in a vat or whatever (I think this is exceedingly unlikely, but still possible), but at least I'm here thinking, whatever that may mean. Dan Lloyd makes a strong case for this being the starting point of any good philosophy in his book Radiant Cool, and I'm basically borrowing that here (as is the case with a lot of this stuff -- some remembered and others not -- see O'Reilly:General_disclaimer).

The primary nature of our inescapable subjective bubble into this world has many important implications. Everything we will ever think, understand, know, believe, imagine, and dream must be contained within our own personal subjective bubble. We project so many things out into the world, without realizing that everything we experience is actually inside our heads. We imagine so much about the world that we don't actually experience. Our conceptions of [other peoples conceptions of] God and Heaven and everything magical are all floating around inside this same little bubble. We are all the undisputed kings and queens of our inner lands. Nobody has the same subjective experience as you (or me), and it is as ineffable and spiritual and magical as all that.

And yet, it is almost undeniably certain that all this magic is the product of a perfectly objective physical system: the brain. I can certainly mess with your experience in a very significant way by altering the electrical or chemical function of your brain. And we equate death with brain death. So, we have a Rosetta stone of sorts: two views of the same thing -- subjective experience, and what we objectively know about the brain. To me, this clearly means that understanding the nature of thought is the most important thing I can do -- it is even more important than physics (so there!). This is (presumably) why I study computational cognitive neuroscience -- how the brain gives rise to cognition (thought), and even consciousness.

There is a Real Objective Outside World

The idea that the world outside my head actually exists and continues to do so in my absence is an assumption that is highly consistent with everything I've experienced. I keep hearing about all this stuff that happened before I was alive. Mostly it is all the same stories. Every time I look up in the sky, the moon and planets are moving according to regular patterns. People replicate various experiments over and over again, with the same results (e.g., measuring the basic electric charge). It is technically possible that this is all a contrivance for my benefit. But it is really very hard to get all those details right again and again. Having tried to simulate artificial realities, it is a very hard problem to get right, and it just seems exceedingly unlikely that this is all here for my benefit alone. If you really think there isn't an objective physical reality, why don't you just will yourself in to flying and doing all sorts of other crazy dreamlike stuff? Hmm. Not so easy. Indeed, evolution has likely pounded into our heads a decent respect for the objective physical world. Those who deny it probably just don't live that long.

I am a Biological System

I am a biological system of extreme complexity, but nothing more (or less -- please, no robot jokes!). This is why I study the brain and don't just invent fanciful mathematical models or other artificial systems for describing how thought might work.

Yes, this means that I must actually confront death in all its finality, but one of the most important lessons I've learned in life is that you have to confront the truth, and eventually embrace the truth whatever it might be. It never works (for very long) to try to avoid it. You can't "get away" with anything, really. Indeed, I have the comfort of knowing that everyone around me is going to meet the same fate as me, even those who subscribe most feverishly to some religion. Just imagine how bad those suicide bombers would feel if it were somehow possible for them to realize that they just aren't going to heaven. Too bad they're just dead and can't know or feel anything anymore, and neither can their unfortunate victims. Just plain stupidity. Same can be said for all the crimes and killing done in the name of all religions.

I am a (no-longer?) Deluded, Denuded Ape

As a biological system, the human brain is a product of evolution, and carries with it a lot of historical baggage. I am a naked ape. It is shocking how hidden this obvious secret is from most people. We are deluded apes as much as we are denuded apes. Somehow most people can live their entire lives and not really confront this obvious fact. We kill, we mate, we eat, we shit, we die. Everyone knows this, at some level, however deeply buried. Children's books are actually quite good at explaining some of this: Everyone poops (except Max of course). This is, coincidentally what chimps and all other apes do. Hey, even rats and snakes and most animals do most of the same stuff we do. We share huge amounts of DNA with rats and other mammals, even fruit flies. Modern drugs can be tested on mice and other mammals, and these results very often tell us what they'll do in people.

The only way I can understand the atrocities that people commit -- every day all around the globe people are murdering, raping, torturing, etc -- is by recognizing that we are apes. I honestly don't know how anyone could watch even just one 20 minute TV news show (or practically any other show on TV for that matter) and not recognize instantly the profound truth that we are just plain old crazy apes! All we care about is sex and food and killing! Again, what a coincidence! That is precisely what chimpanzees and other apes who have been taught how to communicate spend all their time talking about!

Of course, we are also clearly very different from chimpanzees in our ability to think abstractly and communicate with complex language and interact with our rich social culture (see below). But everyone knows this, so I don't really need to dwell on it.

I Think, Therefore I am Wrong (most of the time)

There is another critical lesson from the previous point: just because we think, doesn't mean that what we think has any particular relationship to the truth. Cogito ergo wrongo (most of the time). People can be absolutely blind to the most obvious facts. The idea of the number zero, negative numbers, or any accurate conception of the most basic laws of physics or the relationship of the planets and the sun, took a really long time to figure out. The fact that we are fundamentally apes is still widely underappreciated!! Ok, I said that already.

Our amazing ability to think wrong is actually a direct consequence of our ability to be so smart. It is just another self-consistent piece of the puzzle (see that link for more details).

Here are some important consequences of this point: It takes a lot of effort and determination to figure stuff out. It doesn't come naturally. Most of the time is spent being wrong. Get used to it. Cherish any little scrap of truth, and try to build upon it. Question all of your assumptions -- several of them are likely wrong. Be prepared for your world view to change dramatically -- it is probably wrong. Question reality. Mock Realty. etc.

Of course, the notion that people are generally wrong assumes that there is some way of discerning right from wrong. See above introduction and self-consistency and epistemology for more on that.

The Human Animal is Exceedingly Social

Even the most extreme hermit defines themselves by their social role: one who avoids social contact. I used to think that I was kind of a hermit, a lone ranger in the great sea of nameless ideas and machines intricate and complex. I now know better. I, like all but the most autistic of us, am absolutely driven by and dependent on my social environment. Society shapes what I think, who I think I am, what I want to do with my life, who I like, who I hate, just about everything good and bad about me. Michael Thomasello makes this point very well: what makes us human is our extreme social orientation and desire (and ability) to share with others, combined with a decent amount of labile cortex to soak up the social juices.

Society Makes us Smart

It turns out that billions of generally misguided people fretting about for thousands of years really do add up to a few smart ideas. Fortunately, these good ideas occasionally get carried forward and spread more widely (memes or whatever you want to call them -- culture more broadly speaking), and that leads us right up to this point in human history where we are a lot smarter collectively than any of us is individually. By a long shot. Some people get disaffected by this, but instead we should revel in the benefits of having been born later than all those other people who died without even knowing what a wiki is!

Apes in Society: In-Group Out-Group Dynamics are the Root of all Good/Evil

As everyone knows, evolution is all about survival. Because we are social animals, our survival depends on maintaining our social environment. Unfortunately, resources being limited as they are, not all social groups can survive. Hence, the human apes that survived (i.e., us), are biologically driven to promote the well being of our own social group (the "in-group"), to defend this group against all others (the "out-groups), and where possible, eliminate the out groups from competing for our resources. This dynamic is strikingly clear in chimpanzee behavior. From this simple set of patently obvious facts, the entirety of good and evil derives.

Good is what occurs among members of the in-group. We nurture, care, love, and sacrifice for members of our own group. This has become amazingly apparent to me and my partner Yuko as we have struggled with our son Max's serious medical condition. People have done unbelievably nice things for us. Someone we never met even offered to donate a part of her liver for Max. Truly astounding.

Evil is what occurs between members of different groups. We hate, kill, torture, and rape members of the out-group. In the context of all the good that we experienced with Max, the news brought daily reminders of the equal and opposite evil that is taking place across the globe. I needn't elaborate -- fresh tidings are available every day.

Ironic isn't it how religion, which tries so hard to deny the blatant fact of our ape nature, does such a good job of reinforcing it! Nothing provides a stronger basis for in-group out-group dynamics than religion. How nice can you be to someone once you believe that they will burn in hell forever because they aren't a member of the same in-group as you? Actually, in point of fact, you can often be quite nice, in my experience. Some of Yuko's best friends in college held (and presumably continue to hold) this very same belief about her. Clearly an indication of inconsistent thinking. However, as we all know, this is not always true, and religious group dynamics are holy fuel on the fire of our ape instincts.

Is it possible (or even desirable) to eliminate in-group out-group dynamics from our society? I'm not sure -- this ranks high on my list of major unsolved problems. However, I do know that probably we as a species need to confront the truth and recognize that this is a very strong aspect of our evolutionary baggage, and we can potentially use our prefrontal cortex to apply top-down cognitive control over subcortical brain areas where these instincts lie. It is just a matter of society working hard to make us smarter than we otherwise would be -- this happens all the time, at least to some extent. Speaking of which, here is my pop psychology self-help guide.

Politics

Politics is all about in-group out-group dynamics. The United States is a 2 group society: Democrats and Republicans. I personally sympathize mostly with Democrats. I can't actually see why anyone would be a Republican. Yuko just showed me this sticker: "Republicans for Voldemort" -- exactly! Republicans always seem to want to hold good new ideas back, putting the brakes on the only force at work that is making us smarter: social evolution. They stand for reinforcing entrenched in-groups and out-groups. They are religious nuts.

But wait a second. Republicans are also "realistic" and "pragmatic" and "fiscally responsible". Hey, I am absolutely for all of those things (except maybe the fiscally responsible part -- I mean, I'm for it, just not actually capable of it myself). Republicans confront the truth in a lot of ways that Democrats supposedly don't. That is my main mantra.

So, you see, this is the problem with in-group out-group dynamics. It makes you just hate those stupid out-groupers, even if maybe they are a lot like you in a lot of ways that you might have never considered.

The Principled Middle Ground: A Political Utopia

My own personal favorite political analogy is the learning algorithm that I developed, called Leabra. Leabra combines two different learning mechanisms (which had always been put in opposition in the literature -- group dynamics again): error driven learning and Hebbian learning. Error driven learning is a lot like socially liberal political ideologies: it tries to solve problems by getting each neuron to make a contribution to the solution in proportion to its ability, and it doesn't take advantage of any local "individual" motivation that the neuron might have. Hebbian learning, on the other hand, is all about the individual neuron: it tries to maximize its own correlational "wealth", without regard to more global problems that might be confronting the network. From these greedy local actions, some global benefit does accrue. However, purely Hebbian networks often fail to solve the relevant problems. Greed only goes so far. On the other hand, purely error-driven networks fail for the opposite reason: the problem gets parcelled out into so many pieces, and each individual only does the bare minimum required, and often the job never actually gets done. Clearly, you need to combine the benefits of each of these mechanisms, to produce the optimal learning mechanism.

I think the same lesson applies to politics. It is in fact a key principle of the US constitution: a balance of powers. One needs a balance of different forces, working together in a cooperative overall environment, to achieve the optimal solutions to societies problems. Extremism is (provably!) very bad. And yet, because we are social apes, we are driven to extremism because of group dynamics. Again, the main problem we as a species need to overcome is our in-group out-group dynamics. Ideally, we become one big in-group. Maybe we'll find some aliens (the outer-space kind!) to out-group?

Toward a Science of Politics

The problem with politics is that there is no firm basis for deciding between different approaches to solving social problems. Sociologists might scoff. But, in practice, political discourse is all about in-group-out-group and I'm right and you're wrong, etc. Clearly, we need a better scientific basis for making important social decisions. But, society is so damn complex, it may seem rather hopeless. However, I think a lot more progress could be made by relaxing the crazy pure math obsession of economists, and using extensive computational simulations (like those used in understanding the brain, in the field of computational cognitive neuroscience. Furthermore, just thinking of society as an organism would go a long way toward framing the issues in a productive way.

Physics: Its All Waves

Finally, if you're going to have a complete, self-consistent world view, you need to go all the way down to the most basic stuff of the universe. What are we really made of, ultimately, and what is this universe that we live in? You can (theoretically) go from the above line of reasoning down to basic physical principles by following The Golden Chain.

I have developed an explicit computational theory of fundamental physics, based on analog (real-valued) cellular automata. For a full treatment, see: http://grey.colorado.edu/WELD/index.php/Main_Page

At present, this theory provides what is quite possibly a completely physically accurate model of the electron, and the associated electromagnetic force (and just maybe something critical about the weak force, and maybe even a decent stab at gravitation). Someone, probably Einstein, once said that "if we could just figure out the electron, that would be enough". Maybe, maybe not. Anyway, I'm pretty excited about it.

The key stumbling block for modern physics (before string theory came in and just made a mess of the place) is how to think clearly about quantum physics. Is it particles or waves? Somehow, it is some weird version of both, and nobody can quite understand it. If you fundamentally don't understand something, and the part you do understand for sure doesn't make any sense (if that does!?), that is a breeding ground for crazy ideas. Recent critiques of string theory suggest that nearly the whole field of physics has gone off the deep end. That's OK, though. Perfectly consistent with the fact that we are all apes, and nobody has any direct access to the truth, and the smarter you are, the more likely you are to be wrong (see individual Cogito Ergo Wrongo). It will take a while to work through this obviously difficult problem. When we do, everyone will rejoice and the word will spread. Hopefully.

Anyway, the idea embodied in my model is that it is fundamentally all just waves. But these waves are non-linearly self-coupled, and this gives rise to emergent phenomena, which look to us like particles. This idea has actually been around for a while, under the name of neoclassical electrodynamics. However, the problem is that it is exceptionally difficult to mathematically analyze these non-linearly self-coupled wave systems. Given that such mathematical analysis is the bedrock of physics, not much progress has been made in this area.

My approach has been to just forget about the math (which I couldn't even dream of doing anyway), and code up the thing in a computer simulation and see what happens when you run it. This approach has served me surprisingly well in my day job trying to figure out how the brain works (which is similarly a non-linearly self-coupled dynamic system of interacting neurons).

It turns out that when you take this approach, the equations (ok, there has to be at least some math) look really really simple (to the trained eye, at least). They're all just waves! And the wave equation is just about the simplest kind of equation you can implement that does anything interesting at all, with the simplest kind of Cartesian assumptions about the structure of the universe (Descartes got a lot of stuff right, actually!).

So, it looks to me like the universe is telling us something very important. Something like, "hey, I'm a big analog cellular automaton, you crazy humans!". Of course, because I can't do the math, I have absolutely no idea if this thing will actually produce all of the known physics, and it turns out you need a really, really, really big computer to simulate even a very tiny bit of the universe. So, I'll have to wait a while on that one. Fortunately, the same computers that are good at simulating brains are good at simulating this physics stuff, so I'm hopeful.

Cosmology

Meanwhile, we can contemplate the unknowable. Let's just say for the sake of discussion that the universe can be simulated by an analog cellular automaton running some nice simple wave equations. How big is it? How long has it been around? For my taste, and this is all anyone has to go on here, I think it would be silly to make this nice uniform cellular automaton, which is just a bunch of neatly arranged 3d cubes permeating space, and then just have it end somewhere. What exactly would be beyond the edge? From a pragmatic perspective, edges are a big pain in these systems.. Why bother? So, I think the universe is infinite. Same arguments apply to time. Space and time are infinite. Cool.

What about the big bang? Probably just a local phenomenon: black holes all coalesce and make a mega black hole that then sucks up all matter up to a point and then explodes. Cleans up all the messes and enables a whole new slew of the same stuff to happen, over and over again forever and ever, wherever and ever...