Friday, July 28, 2006

Will Artificial General Intelligence do art?

Webster defines art as "the conscious use of skill and creative imagination especially in the production of aesthetic objects." The first part of the statement seems to hold for AGI, it is likely, almost by definition, that AGI will have consciousness (e.g.; self-awareness), skills and imagination.

In the extended clause, the term objects is also fine, as objects can exist in the physical world or in digital worlds, do not need to be continuously instantiated and do not need to be instantiated at all to be accorded value by AGI. Aesthetic is the most problematic aspect from AGI's view point as aesthetics could be driven rationally, as in by geometric proportions, or emotionally, as in by feelings. AGI will have rationality but maybe not emotion. Emotion is not likely a precondition for intelligence.

How do humans know something is beautiful? It is a subjective judgment, but so can it be subjective (individual, heterogeneous) for AGI. Humans may claim the process of evaluating beauty is subjective too, but it is likely that the process is routine; sensory information comes into the brain and properties (symmetry, color, melody, etc.) are identified and processed, some of which may elicit emotion. Emotion is possibly an after-the-fact signal for more expedient cortical interpretation but not required for perceiving beauty.

Progressing from this simple analysis that AGI could have the concepts of art and aesthetics, how relevant will this be to its priorities? "The conscious use of skill and creative imagination" will likely be the chief activity of AGI, but not for creating stuff, rather for creating ideas, hypotheses and experiments with aesthetics being defined as simplicity, accuracy and economy of solution. AGI art is scientific discovery.

Using machine intelligence for knowledge generation

What will artificial general intelligence (AGI) do with its infinite time and initially great computational resources? Like humans, AGI will be goal directed, though likely in a more focused way, so AGI would be likely to do the same things as humans but in different ways (e.g.; using parallel, distributed, etc. methods) and with different proportions; these activities are:

1) Control matter
2) Solve problems and fulfill needs
3) Create/discover new knowledge

The difference is that AGI will nearly exclusively spend its time in #3 vs. most humans which do the opposite.

Once securing survivability, AGI will need to spend a minimum of time controlling matter. AGI will also need to spend a minimum of time fulfilling needs as biological needs are eradicated and emotional needs are also likely expunged having been translated into goal motivators.

AGI will be concerned with what any next evolution of intelligence would focus on; solving outstanding problems, being creative, discovering anything not previously known or understood such as a grand unifying theory of physics, faster than light travel, the riddles of dark energy and dark matter and controlling the force of gravity.

Given the benefit of dramatically extending knowledge in these ways and the challenges and long road to AGI, more purpose-specific machine intelligences (narrow AIs) should be build to concentrate on singular problems.

Wednesday, July 26, 2006

Enterprise software-as-service kinks

Marc Benioff's proclamation at the AlwaysOn Innovation Summit at Stanford today that the "Business Web is Here" seems a bit unfounded given a) that its not, b) Salesforce.com's legendary outages, c) the company's 50% drop in stock price so this year (vs. a 15% drop in the NASDAQ) and d) a similarly lackluster showing from software-as-service peers RightNow Technologies, NetSuite and others.

Meanwhile, Oracle's market value has quietly doubled in the last four years as the company has had to provide support for the whole market of enterprise software solutions. 94% of Oracle help desk calls involve integration with other applications. Oracle customers' number one business process challenge is other applications: conflicting instances of invoices, databases, etc. on multiple systems.

The challenge for both established (former darling SAP's recent downside earnings surprise, for example) and startup enterprise software companies remains unseating Oracle which is a formidable obstacle for many reasons including the still problematic software-as-service model. Open source and Web 2.0 solutions are distant alternatives given their current lack of scale, reliability, maturity and implementability for reaching serious CIO consideration. Unfortunately, enterprise software continues to be driven by CEO ego wars as opposed to an innovative look at customer needs.

Tuesday, July 18, 2006

The myriad flavors of spaceflight

The best Earth views can be had for $20 million (plus $15 million for a space walk) which is actually getting cheaper since Russia has not changed the price in the five years that they have been offering space tourist visits to the International Space Station. So far, there have been three paying space tourists and more are scheduled including X Prize donor Anousheh Ansari. The rigorous medical examination and training for orbital flight is detailed here.

If one can be content with suborbital flight and waiting a few years, there are more options, the cost is dramatically cheaper and the six month cosmonaut training is obviated. There are many potential providers of suborbital flight, including Space Adventures (2007-2008 timeframe), Incredible Adventures aboard the Rocketplane, and Virgin Galactic via Spaceship Company ships to be built by Burt Rutan's Scaled Composites. The Virgin Galactic suborbital flights are $20,000 and are due to depart later this decade from a $200 million spaceport to be created in New Mexico.

Zero gravity flights are available right now from a variety of service providers including Space Adventures, Incredible Adventures and X Prize Foundation founder Peter Diamandis' Zero Gravity Corporation, who flies 15 parabola rides for $3750 out of Florida, San Diego and San Jose.

It's great that so many people, both astronauts and cosmonauts (400 to date) and private citizens are having an opportunity for space-related experiences. Hopefully this increasing activity will continue to raise interest, support and involvement in space, astronomy and science in general.

Monday, July 17, 2006

A copy by any other name

If I could make copies of myself, I would probably make lots. The goal would be to maximize my self-development, knowledge, skillsets, experiences and contributions. If I have the privilege and capability of making multiple self-copies, then so would others and therefore resource constraints are implied not to exist. It's not clear how many copies might be good, but probably not infinitely many, unless they could scatter throughout the universe. Would you get annoyed by having a lot of yourselves around? Would they be hard to work with, everyone responding the same way to group dynamics?

Would it be desirable for copies to have circumscribed capabilities and consciousness? Would it be moral to do so? Assuming the easier case that all copies have full capability, volition and consciousness, the idea is to be a borg mind. All copies share all the knowledge and all experiences, syncing on a daily (or more likely real-time) basis, somewhat like David Brin's "Kiln People" dups for example. There might be cases in which a copy no longer wants to participate in the borg mind, for example if it had some really different experiences and got pulled into a new memeplex and lifestyle (beneficial or detrimental). In other cases, the copies could take turns going to work and doing the other obligatory activities of life.

Tuesday, July 11, 2006

Aesthetics and science are distinct

Some scientists have been talking about the old fallacy of physicists that the most simple, most elegant solution must be the correct one. Simple, sometimes as in Occam's razor, is an objective assessment and may be appropriate, but elegant, which is a subjective assessment, is not.

Some aesthetic supporters include Mario Livio of Baltimore's Space Telescope Science Institute in his 2000 book, "The Accelerating Universe : Infinite Expansion, the Cosmological Constant, and the Beauty of the Cosmos" who synthesizes the views of a variety of pro-aesthetics scientists and more recently, well-known game-designer Will Wright at a June 2006 talk hosted by the San Francisco Long Now Foundation. Other long supporters of aesthetics continue to populate mathematics and physics.

Pandering to aesthetics is a surprising and disappointing departure for scientists from their more usual rational stance. Whether a solution is aesthetic or not is irrelevant to its main property of being able to solve a problem. In fact, aesthetic value is normally accorded to a scientific solution once the solution's accuracy is established, and is very much in the eye of the beholder. The Aeron chair, as Malcolm Gladwell reminds in his 2005 book "Blink," was ugly until it became an Internet entrepreneur status symbol.

Evaluating scientific solutions in terms of aesthetics is just another indication of how necessary new conceptual paradigms are for many areas of science where new ways to frame the problems could be an important step forward to discoveries, particularly in particle physics, cosmology and artificial intelligence.

Saturday, July 08, 2006

A financial hedge for your house

After proving itself with Weather Futures, the Chicago Merc launched another important innovative product, Housing Futures and Options, earlier this year. These are financial contracts which allow residential home owners to hedge their long real estate exposure against declines in value by taking a short position in one of the housing contracts. The largest market participants currently are foreign and domestic institutional investors and should eventually include speculators as bid-ask spreads shrink and more options contracts become available. The housing contracts are currently offered for the top ten markets in the US as well as in composite.

The housing contracts are pegged to a set of broad-based US nationwide residential real estate indices, the S&P/Case-Shiller Indices, an important instrument which registers the changes in regional home values in a standardized way. The Indices were developed by Wellesley economist Karl E. Case and Yale economist Robert J. Shiller who discussed the idea for housing futures and other finance innovations in his 2000 book, Irrational Exuberance. Here is a graph of data from the S&P/Case-Shiller indices, showing the general run-up in all residential housing in the last twenty years with San Francisco's Internet bubble sticking up and both San Francisco and New York lagging the national composite in the last two years.

The aggregate value of US residential real estate is approximately $20 Trillion (assuming 75% of total consumer debt is mortgage debt and aggregate loan-to-value is 45%) meaning that it is one of the largest asset classes which has previously not had highly liquid widely tradable standardized financial instruments. Putting the size of the real estate asset class in context, the 2005 US GDP was 12 Trillion, the total US stock market capitalization is $13 Trillion, and the US bond market size is $26 Trillion.

Hopefully there will be an acceleration of new financial instruments from trading exchanges like the Merc and the CBOE and startups like derivatives innovator Hedge Street and consumer loan community Prosper.

Monday, July 03, 2006

Is your city a Walking Detroit?

Outside of the ten largest cities, the Creative Class cities and the "new" sunbelt cities (Las Vegas, Phoenix, Denver, Ft. Myers, etc.), US cities may be little other than "Walking Detroits," cities whose populations and economies are in ever greater stages of languishment. Pennsylvania and upstate New York is a living museum of Walking Detroits: Scranton, Harrisburg, Lancaster, Syracuse, Albany, Rochester, Buffalo, Binghamton, etc. There is a palpable sense of the decaying physcial infrastructure being too big for the inhabiting population and that the heyday is long and irreparably gone. Paul Graham and Richard Florida have the best turnaround suggestions.

The interesting point is that these cities were the vibrant centers of the country a hundred years ago. In one hundred years, where will the vibrant centers of the nation or region or world be? It is easy to see where they are now but it is unlikely that they will be static for the next hundred years. What could trigger the decline of today's teaming centers?

Civilization was previously much more limited by natural resources, transportation and the more manual aspects of life. The biggest shift in the last hundred years has been that proximity to natural resource extraction locations and transportation matters much less. There is now more emigration and people are propelled to centers of thought innovation (which are synonymous with economic and other opportunities). Jared Diamond notes in Collapse that the key means to a civilization's survival is sustaining natural resources. Perhaps the geographic pendulum will swing back to desired proximity to natural resources (this time water, clean air and naturally grown foodstuffs) as metaverse worlds continue to replace physical world proximity requirements.