Taner Edis

How Gödel‘s Theorem Supports the Possibility of Machine Intelligence

Gödels Theorem is often used in arguments against machine intelligence, suggesting humans are not bound by the rules of any formal system. However, Gödelian arguments can be used to support AI, provided we extend our notion of computation to include devices incorporating random number generators. A complete description scheme can be given for integer functions, by which nonalgorithmic functions are shown to be partly random. Not being restricted to algorithms can be accounted for by the availability of an arbitrary random function. Humans, then, might not be rule-bound, but Gödelian arguments also suggest how the relevant sort of nonalgorithmicity may be trivially made available to machines.

The Rahmi M. Koc Museum, Istanbul

The past decade has witnessed the establishment, rapid growth, and gradual maturing of the Rahmi M. Koç Museum in Istanbul, Turkey’s only museum of science, technology, and the history of technology. Its focus on transportation, industrial, and communications technologies is familiar to European and American visitors, but new for Turkish citizens. While the Koç Museum displays objects from all over Europe and the United States, it seeks to highlight Turkish perspectives on history of technology. A small but dedicated staff has significantly expanded and improved the exhibits over just the last two years. Museum funding comes entirely from the Rahmi M. Koç Museum and Cultural Foundation, whose head, Rahmi Koç, chairs the Koç Group, Turkey’s largest corporation. In 1928 Rahmi’s father Vehbi became Turkey’s representative for Ford Motor Company and Standard Oil. Over subsequent decades, Vehbi set up Turkish manufacturing arrangements with Siemens, Fiat, and General Electric. In the 1960s, Koç companies began manufacturing the Anadol, the first car completely built in Turkey; the Koç Group also produced the first locally manufactured examples of many domestic and industrial items, from light bulbs to tractors. Describing the museum’s origin, Rahmi Koç explains that since childhood he has been fascinated by Istanbul’s steam locomotives and ferries as well as imported model trains and clockwork toys. Visiting England, Koç writes, he “saw how much importance they attached to heritage and to the machines and tools from the Industrial Revolution.” Entering the family Ford agency and preparing to begin Turkish auto manufacture, Koç trav-

Unusual Constraints in the Quantum Statistical Mechanics of Josephson Junction Systems

In order to apply quantum statistical mechanics to systems composed of Josephson junctions, the unconventional constraint of fixed “macroscopic wave function” magnitudes on either side of a junction must be accommodated. In order to use this information, the density matrix formalism must be extended to deal directly with probability distributions over general quantum states. As a result, in thermal equilibrium, the explicit temperature dependence becomes modified from the trivial 1/kT factors.

Technological Progress and Pious Modernity: Secular Liberals Fall Behind the Times

Since the European Enlightenment, secular liberals have imagined themselves to be at the forefront of technological and social progress. Our celebration of improved human prospects, however, is often laced with ambivalence: we have acquired greater power to destroy one another and the natural environment, and to produce more inventive forms of oppression. Therefore, secular liberals have often been accused of being too optimistic about the technical progress that shapes modernity. Today’s political climate, however, undermines secular notions of progress in a different way. The continued dominance of business-class conservatism, often allied with religious movements, has given secular liberal aspirations a nostalgic edge. Conservative constituencies have a better claim to be shaping technological progress today. It has become increasingly doubtful whether secular liberals represent the cutting edge of modernity any more.

Rejecting Materialism: Responses to Modern Science in the Muslim Middle East

In the past centuries, most Muslims have encountered modern science as a Western import. To avoid being overwhelmed by the military and commercial advantages enjoyed by technologically advanced nations, Middle Eastern Muslim societies had to begin adopting modern knowledge. As westernization started to shape social structures and institutions as well as technologies, conservative Muslim responses to modern science typically became conditioned by the demands of cultural defense. Many Muslim thinkers argued that upholding the religious character of Muslim civilization meant borrowing technology but rejecting the perceived materialism pervading the conceptual frameworks of modern science. This defensive approach remains prominent in present Muslim thinking about science. Almost all religiously oriented Muslim thinkers take harmony between science and Islam for granted, but in practice, conservative Muslims often express deep reservations about the naturalistic perspectives dominating modern science. Especially in the popular literature, religiously motivated distortions of science are common. Darwinian evolution is a particular target of rejection.

Stratospheric Relaxation in IMPACT's Radiation Code

While Impact incorporates diagnostic radiation routines from our work in previous years, it has not previously included the stratospheric relaxation required for forcing calculations. We have now implemented the necessary changes for stratospheric relaxation, tested its stability, and compared the results with stratosphere temperatures obtained from CAM3 met data. The relaxation results in stable temperature profiles in the stratosphere, which is encouraging for use in forcing calculations. It does, however, produce a cooling bias when compared to CAM3, which appears to be due to differences in radiation calculations rather than the interactive treatment of ozone. The cause of this bias is unclear as yet, but seems to be systematic and hence cancels out when differences are taken relative to a control simulation.

Update on the Radiation Code in IMPACT: Clouds, Heating Rates, and Comparisons

This is a summary of work done over two months in the summer of 2005, which was devoted to improving the radiation code of IMPACT, the LLNL 3D global atmospheric chemistry and aerosol model. Most of the work concerned the addition and testing of new cloud optical property routines designed to work with CAM3 meteorological data, and the comparison of CAM3 with the results of IMPACT runs using meteorological data from CAM3 and MACCM3. Additional related work done in the course of these main tasks will be described as necessary.

Testing Impact?s Radiation Code

This is a summary of work done over an 8 week period from May to July 2004, which concerned testing the longwave and shortwave radiation packages in Impact. The radiation code was initially developed primarily by Keith Grant in the context of LLNLs 2D model, and was added to Impact over the last few summers. While the radiation code had been tested and also used in some aerosol-related calculations, its 3D form in Impact had not been validated with comparisons to satellite data. Along with such comparisons, our work described here was also motivated by the need to validate the radiation code for use in the SciDAC consortium project. This involved getting the radiation code working with CAM/WACCM met data, and setting the stage for comparing CAM/WACCM radiation output with Impact results.

Weak-link magnetically modulated resistance response in granular superconducting systems.

Experimentally, magnetically modulated resistance (MMR) techniques are effective in ascertaining the properties of granular superconductors, which can be modeled as a network of Josephson junctions. Each junction is parametrized by a critical current, with a capacitance and resistance in parallel to account for the nonsupercurrent components. The overall current-voltage behavior of the network, in the presence of an external magnetic field, is given by a set of coupled nonlinear ordinary differential equations, which are studied numerically. Features of interest in the simulated experiments arise from effects of the applied field on couplings, and the transitions between multiple solution classes of the differential equations. Through these simulations of the phase dynamics, details of the weak link MMR signal can be understood, the typical signal being reproduced with large enough networks.