Supercomputers surpass 2.5 petaflops

The biannual list of the world's fastest supercomputers was released on November 13, 2010. For the first time, supercomputing capability surpassed 2.5 petaflops with the world's fastest supercomputer, the Tianhe-1A - NUDT TH MPP, X5670 2.93Ghz 6C, NVIDIA GPU, FT-1000 8C NUDT, at the National Supercomputing Center in Tianjin China, clocking in at over 2.5 petaflops.

Figure 1 illustrates how supercomputing power has been growing in the last five years, starting at (a paltry) 136.8 gigaflops in June 2005, and experiencing four solid doublings in growth. This rate of progress is estimated to continue, and usher in the exaflop era of supercomputing by mid-decade. The IBM Roadrunner at Los Alamos was the first to achieve speeds over one petaflop in June 2008 and held onto the fastest computer seat for three measurement periods, then was surpassed by the Cray Jaguar at Oakridge for two measurement periods. China has now captured the fastest supercomputer ranking with its NUDT MPP.

Figure 1. Growth in Supercomputing Capability: Jun 2005 - Nov 2010

Source: Top 500 Supercomputing Sites List Releases

The world's supercomputers are working on many challenging problems in areas such as physics, energy, and climate modeling. A natural question arises as to how soon human neural simulation may be conducted with supercomputers. It is a challenging problem since neural activity has a different architecture than supercomputing activity. Signal transmission is different in biological systems, with a variety of parameters such as context and continuum determining the quality and quantity of signals. Distributed computing systems might be better geared to processing problems in a fashion similar to that of the brain. The largest current project in distributed computing, Stanford protein Folding@home, reached 5 petaflops in computing capacity in early 2009, just as supercomputers were reaching 1 petaflop. The network continues to focus on modeling protein folding but could eventually be extended to other problem spaces.

Questions uplevel the web

A new internet trend has popped up – asking questions to friends and strangers. Questions are a light-weight means of bringing structure and context to social interaction.

Many people start off with a low level of engagement (e.g.; what is the best pizza place in your neighborhood?). Things can start scaling up (e.g.; how much does it cost for a family to live in London on average?) and become quite detailed (e.g.; can a publisher see the average CPC or eCPM of a particular ad or advertiser on AdSense?). More importantly, more complex thoughts, opinion, philosophy, and value systems are being elicited (e.g.; what is a good real-world example of the "prisoner's dilemma" in recent history?).

The question-asking trend began with Hunch asking users questions to draw a taste profile map for users. The current generation of the concept features sites like Quora and formspring where users ask each other questions. Users have profiles and friending/following capability, with their questions and answer activity appearing in a status feed. It is like a friend network with topic precision.

Metrics would be interesting, such as the level of semantic complexity of questions, the number of repeated questions (an indicator of what is on the collective mind), and the ratio of asked to answered questions.

Question networks could be the real start of the semantic web, the natural automated approach (akin to Google’s automated machine learning) for creating the next higher level of information density.

Ultimate possibilities for life and technology

Thinking really long term, what would it be like if all matter, including life, could be designed and built to specification with nanotechnology and synthetic biology? Form factor could become ephemeral and purpose-driven. An intelligence could embody as a human, as a fleet of starships, as a crane, as a school of nanoparticles, or remain digital.

Some interesting issues could come up, say from having multiple persistent copies of one intelligence. What would the social, legal, economic etiquette and governing laws be? Or would these words even make sense anymore? The notion of the distinct individual may become obsolete.

Transhumanism will be an interesting and certainly divisive step, when groups or all humans have radically enhanced capabilities as compared with today. Posthumanism, the moment of speciation, may be quite a shock.

What about utility functions? In a digital format, traditional biological functions make a lot less sense. And what about emotion? Is there a relevant adaptation for the digital substrate or is emotion just another biology-based information system?

What is intelligence and is it reflected differently in a digital medium without the sensory input context of the physical world? Maybe intelligence is nothing more than manipulating patterns of information.

Finally, what are the ultimate possibilities for life and technology once joined? What if any would the activity be? Would the focus be on aesthetics? Analytics?

Heard from the future…

• I don’t feel like myself in this upload anymore
• I only got 98% fidelity in my revivification
• You must be a newbie, you never get full reload on a reviv
  
• I don’t want to be so present to my experience
• I don’t want to remember all my memories, I want to edit out x, y, z and enhance a, b, c and…
• I don’t want all my actions (and possibly thoughts) recorded by public cameras and sensor nodes
• Why did I reembody as a (man, dog, robot, hummingbird, bee, network node, airplane, etc.)?
• What is the construct "I"?
• Is this virtual world virtual? Where is the real virtual world?
• I think that AI over there is a few methods short of a class library

Technology intervention is moral

Advanced civilizations may have policies for interacting with civilizations deemed to be less advanced than their own. On Earth, there is currently no cohesive national or global view on contact with any non-Earth based intelligent societies. In the case of Earth-based societies, interventionism has been the norm.

Assuming safe interaction and communication can occur and intelligence or proto-intelligence has been established,

it is the moral obligation of any more advanced society to interact with any less advanced society.

It is a moral obligation to intervene for the purpose of technology-sharing first and most importantly to ameliorate suffering and improve quality of life, consider vaccines for example. Second, it is patronizing to decide whether or not to expose the less advanced society to more advanced technologies. The moral and respectful path is to expose the newtech and let the other decide.

Third, a broad goal of humanity is to lift all intelligent beings to an optimized state of fulfillment and contribution, so absent existential risk to the more advanced civilization, there is no reason not to share technology. Fourth, considering the 'do unto others' principle, the majority of humans would likely support intervention.

An alternate but less tenable view is that intervention is immoral, that the independence of the other civilization should be respected. The more advanced society does not have the right to interfere. It is better to let someone learn for themselves instead of teaching them; forget the matches and wait a few more centuries for lightning to zap the meat. However, even if the intervention is resented later, it is still more moral to intervene in the sense of improving suffering, quality of life, etc.

Alt approaches to AGI

50+ year old attempts at creating AGI have not been successful. It is possible that AGI cannot be generated from current methods and technologies; the wrong tool is being used, sort of like trying to build a 747 with a toothbrush. Electromagnetism, silicon and Von Neumann architectures may not ever have the capacity to achieve AGI even allowing for continued increases in processing, storage and memory and architectural shifts such as parallelism.

Other substrates might work
Getting around the rigidity of Von Neumann, mathematical, logic-based, computational approaches, symbolic approaches and traditional computers, other computational substrates like quantum computing, DNA computing, etc. might work and also those that humans have not yet invented, discovered or exploited for this purpose like light, air, memes and information. There must be other substrates, and other viable approaches that are not constrained by mathematics and logic.

Information as a substrate
Narrowly, the only existing example of general intelligence is the human brain and the basic requirements of AGI are self-replication and self-improvement. Considering self-replication, there are many examples of more effective self-replication than humans, for example, memes, disease and microbes. Considering self-improvement, memes also self-improve more effectively than humans as they are refined through repetition, and have the unbounded ceiling for improvement of true AGI.

Taking advantage of the self-reproducing and self-improving properties and using memes and information as a novel computing substrate might be one way of extending AGI progress.

Information as a substrate could be developed symbiotically with a very broadly applicable new understanding of the laws of physics based on information and entropy as opposed to mass and energy.

Machine creativity

What are the relevant differences between humans and the machines of the future?

The general claim is that the uniqueness of humans is creativity, imagination and fallibility.

Creativity is rationality
Examining and demystifying the concept of creativity suggests that machines can be every bit as creative as humans. Creativity is merely a new idea or approach, a novel solution to a problem, a fresh representation; a process, a personality attribute, a mindset, an approach to life.

Alternative intelligences/machines can use brute force to rationally crunch through the set of all possible answers to a problem and suggest which are best. Much human creativity comes from "out-of-the-box" thinking which is largely applying knowledge, structure or skills from another domain, and also making mistakes (penicillin, 3M's post-its, Nike's waffle soles, painter Apelles' foam depiction, etc.). Machines can easily do all of this and more, testing a wide range of "out-of-the-box" domains and applying inverse or orthogonal analysis to incorporate human creativity by trial and error.

It is not clear that humans have any positive aspect that cannot be replicated or superseded by an alternative intelligence/machine. Therefore, nothing appears to be lost in the potential extinguishment of the human form as intelligence evolves to non-biological substrates.