I was struck that of the “important ethical issues” Henry Markram refers to regarding building a completely simulated human brain in “The Human Brain Project,” the only one he raises is that of a superintelligent nemesis being created. He does not appear to consider the ethical obligations we would have toward the mind we had created. I worry about the precarious humanity of the minds we would create and about the humanity of the researchers who could, with the touch of a button, give a being with memories and an expectation of the future—if this all works as Markram hopes it will—autism, schizophrenia or a progressively degenerative disease. Who will turn off the simulation when the virtual mind begs them not to?
Robert A. Rushing
University of Illinois
Markram glosses over the key potential benefit of the project: understanding the human brain may allow us to augment intelligence and eventually create superintelligent nonbiological humans.
It also raises a key metaphysical question: If the simulation of the human brain is deterministic, how can it have free will? Or is it impossible to fully simulate human cognition on a deterministic machine?
Avishay Gal-Yam's “Super Supernovae” discusses how stars once thought to be too massive to explode have resulted in supernovae more powerful and longer lasting than any previously observed.
Gal-Yam describes how the production of electrons and positrons removed such stars' supporting pressure of gamma rays, leading to their sudden collapse. But he didn't say what happens to the positrons. Wouldn't they collide with the electrons, be converted back to gamma rays and thus restore support for the star?
Also, do gamma rays from a positron-electron reaction have a characteristic wavelength that can be observed?
GAL-YAM REPLIES: Indeed, positrons produced in the hot core of the star will eventually annihilate with electrons and produce pairs of gamma rays with a particular energy equal approximately to the rest mass of the electrons. This process takes time, however, which means that at any given point, the energy (that was previously completely carried by photons) will now be distributed between photons (which provide pressure) and electrons and positrons with energy dominated by rest mass (which do not). Thus, the overall pressure drops, the core contracts, and so on.
The gamma rays produced by electron-positron annihilation do have a characteristic energy (511,000 electron volts, or 511 KeV), but they are unobservable because the envelope of the expanding star is not transparent; the gamma rays interact with electrons and ions in the outer layers of the expanding, exploding star and are converted to lower-energy photons, which we eventually can observe as light.
PUBLISH AND PERISH?
“Waiting to Explode,” by Fred Guterl, addresses the controversy over publishing two recent studies on the development of H5N1 flu strains that are transmissible among mammals (both studies have since been released). As a scientist, I initially felt it was necessary for all the H5N1 bird flu results to be released: publications allow other scientists to continue projects, and researchers have a responsibility to communicate their data to other scientists. After careful consideration, however, I now feel that submitting all the data was a mistake. The results from this project could help terrorists perfect an airborne delivery system to infect humans.
The solution to this problem would have been to publish some of the scientific findings but restrict the key elements—namely, precisely how to make changes to the viruses that would create an airborne entity. These undisclosed methods could have been shared on a case-by-case basis among researchers, which would have allowed for the continued examination of data among responsible parties trying to enhance public health.