[ExI] How should mirror life research be restricted?

[John Clark]

Ben Zaiboc wrote:

*> I'm confident that nobody can survive an actively hostile ASI. This dumb
> human can think of several reliable ways of ending all human life, given
> the intelligence and resources to implement them, so I'm quite sure Mr.
> Hostile Superintelligence can think of more, and better, ones*


*You are correct. **The September 15, 2025 the Journal Nature contained
this article: *


*How should ‘mirror life’ research be restricted?
<https://www.nature.com/articles/d41586-025-02902-2?utm_source=Live+Audience&utm_campaign=b2c87a0562-nature-briefing-daily-20250916&utm_medium=email&utm_term=0_-33f35e09ea-50169436>*

*The article doesn't mention the enormous increase in intelligence we've
seen in AIs over the last couple of years but I think it adds more urgency
to the question. And I thought the following quotation was especially
interesting: *

*"Over the past decade, molecular biologist Ting Zhu at Westlake University
in Hangzhou, China, has been constructing systems to ease the production of
large mirror-image biomolecules. He started with mirror-image polymerases —
enzymes that can copy mirror-image DNA and transfer that genetic
information into mirror-image RNA. In 2022, he used that approach to make
key building blocks of a simplified mirror-image ribosome,  the complex
cellular machinery that translates genetic information to build proteins. A
mirror-image ribosome, Zhu says, “could dramatically accelerate
pharmaceutical discovery by enabling high-throughput production of
mirror-image peptides. **Zhu, who cannot attend the Manchester meeting,
says he has never wanted to create a mirror-image cell, and adds that he’s
still many years away from even making a functional mirror-image ribosome.
There are some who think his work on the mirror-image ribosome could pose a
threat, however.  John Glass, a synthetic biologist at the J. Craig Venter
Institute in La Jolla, California says “It is my view that achieving
construction of a mirror ribosome is probably the hardest part of the
process of making a living mirror cell” . So, is he calling on Zhu to halt
his work on the mirror-image ribosome? After a long pause, Glass answers:
“Maybe. I think it shouldn’t be made.”"*


*A nuclear war would kill billions of people and destroy civilization but
it wouldn't cause the extinction of the human race, it wouldn't kill every
last person, however something else could. When an AI develops
Superintelligence it will have the ability to create "Mirror
Life", synthetic organisms built using mirror-image versions of the
molecules that make up natural life, they would be composed of mirror
proteins and mirror nucleic acids. The immune system of animals wouldn't
even be able to see mirror life, much less attack it. And not just animals,
the same would be true of plants. Nothing would be safe from it, not even
viruses.*

*You might object and say that if the immune system can't attack it then it
can't attack us, but they could steal molecules in the environment that are
vital to us such as glycerol, because they are achiral (they do not have
mirrored forms), and thus could be consumed by mirror bacteria. And
mirrored cells would still be able to harvest nitrogen from ammonia,
nitrate, and other achiral organic compounds. Escherichia coli is the most
common bacteria in the human body and it can grow robustly in an
environment without chiral nutrients**, and so would its mirror life
alternative. Also, several Nobel Prize winning scientists have pointed out
that it would be possible to genetically engineer mirror life so that it
could metabolize abundant chiral molecules such as D-glucose, common table
sugar.   *


*Confronting risks of mirror life
<https://www.science.org/doi/10.1126/science.ads9158>*

*John K Clark*
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20251005/d9a3db4f/attachment.htm>