Re: Jamais Cascio's Critique [Was RE: Recipe for Destruction - Joy/Kurzweil NYTimes Op-Ed]

From: Eliezer S. Yudkowsky (
Date: Wed Oct 19 2005 - 10:58:05 MDT

Tyler Emerson wrote:
> Jamais Cascio of has written one of the better online
> critiques of "Recipe for Destruction" seen to date:

Okay, I'm persuaded. I had no idea that the published gene sequences
were that useful - I thought it was just sheer principle on the
journal's part. (The researchers, according to Kurzweil and Joy, did
not want to publish the gene sequence to open-access databases, but were
forced by the journal to do so on pain of their paper being refused
publication - not that I regard this as a very impressive threat, but
maybe things are different in academia.) And there are other deadly
gene sequences out there already, such as, if I recall correctly,
smallpox; so publication of H1N1 would only pose a small incremental risk.

I am not comfortable with the overall situation. Perhaps someone should
monitor the protein synthesis / peptide sequencing machines that could
actually reconstruct the flu, in addition to a Manhattan Project for
technologies that promise rapid cures for new viruses, Kurzweil's bet
being on RNA interference (and it seems like a good one).

Actually, brief Googling says that smallpox is a huge virus, not readily
subject to artificial synthesis (like the short virus polio, which was
artificially synthesized by one group of highly trained researchers),
and I'm not sure if other short viruses such as Ebola are publicly
sequenced. The argument that we have existing immunity to H1N1 strikes
me as weak, but I am not an immunologist. Taking this into account, I'm
tentatively persuaded to Cascio's point, but I'm not very persuaded.

I reiterate that the end of scientific openness is not the end of the
world. I used to think that way, until I read Rhodes's "The Making of
the Atomic Bomb" and Lanouette's "Genius in the Shadows", and learned
that this had happened before, and that scientific secrecy had been
successfully argued and successfully implemented in an iffy-seeming
early case, and that the results were, if not utopian, then probably
better than unrestricted openness would have led to.

Eliezer S. Yudkowsky                
Research Fellow, Singularity Institute for Artificial Intelligence

This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:52 MDT