Spivack & Kelly Pushing Tech / Consciousness Boundaries, But How Deep is the Rabbit Hole?
November 05 2008 / by Alvis Brigis
Category: Technology Year: 2008 Rating: 4 Hot
“The web is going to wake up. It is already awake because we are awake and we are a part of it.” – Nova Spivack, Singularity Summit 2008
With their recent blogologue concerning the evolution of consciousness, Kevin Kelly of Wired fame and Nova Spivack, creator of Twine, are spearheading a shift away from the commonly held view of a future in which Strong AI grows in a box, to one in which the Cloud or the Planet is the box. Both are striving to broaden the context in which terms like technology, information, intelligence, communication and consciousness are defined. This is a very necessary step as most of the recent theory and development has been dominated by reductionist AI and technology thinkers who seem to view such phenomena in a vacuum.
Clearly, technology, information, intelligence and consciousness (TIICC) do not exist in a vacuum. In his latest post, Kelly expands his definition of the emerging Technium to include the concept of meta-system transition (advanced by Turchin and Heylighen) that Spivack advocates. Thus, both are now in agreement that TIICC are dependent on the system, which is a very positive development, but also brings them out onto a slippery memeslope.
Because there is no such thing as a closed system (as Godel taught us), it is near-impossible, or perhaps fundamentally impossible, to create functional, highly-useful definitions of TIICC. Kelly and Spivack both concur with this reality:
Kelly: Sadly, there is no ironclad definition for some of the terms we most care about, such as life, mind, intelligence and consciousness.
Spivack: We really don’t know what space, time and energy really are. ... Space, time and energy are inferred by effects we observe on material things that we can measure. I think the same may be true of consciousness.
Though on the one hand we can argue that both meta-systems and certain domains are intuitively distinct and therefore we can form testable definitions for TIICC, on the other we can also make the case for concepts such as meta-meta-systems or super-systems that transcend defined meta-systems.
These may include ideas such as Lovelock’s Gaia hypothesis (both Weak and Strong versions), de Chardin’s Omega Point, Bostrom’s Simulation Argument and John Smart’s robust Evo Devo Universe Model that potentially supports broad cosmic consciousness.
So where does it end? Upon which boundaries can we prop up our definitions?
Our concepts of terms like “TIICC”, “humanity”, “real”, “virtual”, “digital” and even “life” itself, are all contingent on our awareness and concept of the broader system(s) in which we exist. Thus, the only hope for true computational closure may be as far off as the Total Quantification of the entire system (something that Godel would likely argue is impossible until all system resources are devoted solely to this self-simulation), the pursuit of which appears to be a natural Evo Devo drive or property of life and our complex system.
As we compulsively and collectively strive for such quantification, we contemplate just how large, complex and multi-dimensional the Global Brain (Heylighen & Russel), Global Body, universe and cosmos could be.
From a subjective viewpoint, we accordingly develop increasingly advanced systems of memes, math, space-time models, science, etc, to help us achieve better macro-sight and micro-sight of our total system. These help us to convert more information into more knowledge which we then network together to generate more capability (lots of little and big feedback loops). We steadily get better at abstraction and STEM Compression, as Smart puts it, of processes, information and perhaps intelligence, life, and our cosmic fabric.
But still, try as we do, we can’t crack some of these fundamental definitions. Even Smart, a leading cosmic and technology philosopher and good friend, acknowledges how:
We cannot yet know whether ‘IC’, information and its computational emergents, including intelligence and consciousness, can be fully described as simply a special set of arrangements of universal STEM (Space, Time, Energy, Matter), or whether they are also something as or more ‘basic’ and ‘real’ than the physical universe they coexist with.
Of course, the problem resides in the incompleteness and/or inter-connectedness of the domains we are looking to define. None can be defined without broader context or the others. Still, we need/seek to better define them in order to generate better context. It’s a co-dependent, chicken-or-the-egg situation.
But perhaps by realizing that our system seeks to learn and grow by Quantifying itself (recursive processes) we can begin to look for some tendencies, distributions or power laws that can help us to better our understanding of how TIICC work, a process that requires more unconventional experimentation.
For example, if we know that a concept like IQ is system dependent, as cognitive historian James Flynn has shown, then we can work to systematically peg it to the broadest and most complex systems definitions we have available.
Whereas Flynn says that IQ tests like the comprehensive Wechsler Intelligence Scale for Children “should be supplemented with tests of critical acumen”, including “social analysis and criticism, that is, whether they have learned to use concepts like market, tautology, placebo, to analyze what they hear and read”, it could be further supplemented by placing the agent in broader systems context and then measuring to see how little STEM and information (STEM-I as Smart puts it) they require to achieve stated measurable goals, which can also be measured by STEM and informational $ value. This seems like it would be more in line with what people consider to be common sense and would also further widen the concept of intelligence to include the network and elements such as Google, other people, and technology.
Similarly, if we know that a concept like information is system dependent, then we can expand Shannon’s notion of Channel Capacity (the max amount of digital data that can be transmitted over a digital pipeline) to include the information systems or networks on either end of such pipes. We can devise crude tests measuring, for example, the maximum STEM-I effect that a human winking at another human can have. By setting up different game contexts where winks at the proper times result in certain value, we can begin to add systems context and other measures of value to determine maximum information transfer figures.
Though conducting such contextual experiments will certainly not lead to nice, neat, closed-off definitions for terms like TIICC in the short-term, they may help to reveal the extent to which these systems are interconnected and just how much force certain domain components can generally or maximally exert on their surrounding “co-systems”. By avoiding limiting generalisms of such terms and throwing more resources at pattern recognition and exploration of their co-dependencies and growth patterns, we can perhaps collectively create the conditions for the emergence of new terms and frameworks, especially if we find that systems and meta-systems tend to behave in similar ways on many different levels, something Evo Devo thinkers call scale invariance.
This will be difficult so long as we continue to cling to the familiar memes and memeplexes that we tend to rely on, though they have greatly helped get us to this point in our system.
Therefore, I urge thinkers like Kelly and Spivack to 1) constantly question their fundamental definitions and assumptions regarding the terms they seek to define and also use to build up worldviews and 2) to reach out to the budding Evo Devo community that is coming at these same problems from a macro-philosophical, yet observational-by-domain, perspective.
My gut tells me that much valuable memetic progress will be made as Kelly’s first-person cultural perspective, Spivack’s hands-on Buddhist framework, and Smart’s super-structured systems theories intermingle to further drive our simulation of the now and near-future just as we enter the Kurzweilian acceleration era. ;)
Conclusion: It appears to be no accident that the swing from reductionism to holism is underway now that the domains we call information, communication, technology, and perhaps even intelligence are converging and co-evolving at an accelerating clip. As we attempt to better define and quantify such vmeta-systems and/or properties, it will be helpful to rely on both the Global Brain, which contains emerging theories such as Evo Devo and brains, and the Global Body, which includes other technologies, to help us realize just how far both extend, and just how deep our rabbit hole goes.
Just how smart is the system? How many meta-systems does it contain? How do biological, geological, technological, human, universal and cosmic systems co-compute, expand and evolve? Are we effectively living in a “digital” chain of simulations? These are questions we must answer in order to develop the memes and technologies we need to more effectively understand why we are compelled to answer them. :)
Comment Thread (2 Responses)
-
Wow Alvis. I think you broke or blew my mind more than once in this post. I can’t wait to learn even more about the concepts. Here is my two cents:
1. All intelligence is only as useful (or intelligent) as the effect it has on it’s surroundings. Consequently, people and/or machines have to grow and evolve together.
2. I agree that theories of strong AI and reductionism seem to be giving way to more cloud/planet Global mind theories along with augmented intelligence theories as we become more and more interconnected through communication. It’s exciting to think of this in terms of a broader systems evolution or evo devo process.
Posted by: Mielle Sullivan November 06, 2008
Vote for this comment - Recommend -
Wow. I’ve just spent the last couple of weeks researching related subjects, specifically the idea of scale-invariance, and you’ve given me a boatload of meme references I had no idea existed. Excellent work.
My position on the Kelly/Spivack discussion is that trying to measure our own superorganism using smartness is ultimately a dead-end. As you outline very eloquently above, our definitions for TIICC are slippery at best. I believe that our observations and measurements are limited to our own scale; we cannot “zoom-out”.
Sociobiologist E.O. Wilson defines a superorganism (in his recent book of the same name) as a measure of social complexity, not “intelligence” of the colony.
Taking an approach based on social behaviour is not only more practical and measurable, it’s also more honest about the ideological implications of joining a Global Brain.
My full response can be found here, at my blog The Connective.
P.S. Great posts by the way. Glad I found your site. You’re going straight on the blogroll, Alvis. Would love to know what you think of my blog too.
Posted by: ESivan January 05, 2009
Vote for this comment - Recommend