It’s the “barring other physical limits” that is the important caveat. I completely disagree with the statement that labor is the factor that makes CDR (and everything else) expensive.
From: geoengineering@googlegroups.com [mailto:geoengineering@googlegroups.com] On Behalf Of Michael Trachtenberg Sent: Wednesday, September 07, 2016 9:23 AM To: adamd...@ucla.edu Cc: R. T. Pierrehumbert <phys1...@nexus.ox.ac.uk>; Greg Rau <gh...@sbcglobal.net>; bmer...@mercerenvironment.net; andrew.lock...@gmail.com; geoengineering <geoengineering@googlegroups.com>; Andrew Revkin <rev...@gmail.com>; cla...@onid.orst.edu; Oliver Morton <olivermor...@economist.com>; Oliver Morton <omeconom...@gmail.com> Subject: Re: [geo] Scientists Focused on Geoengineering Challenge the Inevitability of Multi-Millennial Global Warming While I agree with your projection please note nothing is costless. A major reason chip labor is used is that it is still cheaper than AI/robotics. As that changes, for better or worse, unemployment follows. Cost less NOT costless. Michael Trachtenberg, PhD Visiting Scientist Department of Chemistry and Chemical Biology Rutgers University New Brunswick, NJ mi...@aesop.rutgers.edu <mailto:mi...@aesop.rutgers.edu> 609-610-6227 On Sep 6, 2016, at 10:52 PM, Adam Dorr <adamd...@ucla.edu <mailto:adamd...@ucla.edu> > wrote: The connection is that economic cost is the CDR feasibility bottleneck. And barring other physical limits, labor is the factor of production that makes CDR (and everything else) expensive. Machine labor obviates this feasibility bottleneck. Choose any product or service, trace its supply chain to its origins, and this becomes obvious: a functionally unlimited supply of costless skilled labor straightforwardly renders all commodified goods and services superabundant (though obviously not infinite). CDR is not feasible today because it would cost trillions of dollars to build the tens of thousands of building-sized direct air CO2 capture facilities and storage needed to draw 5+ Gt of carbon out of the atmosphere. And the reason why it would cost trillions of dollars is because, today, people would have to build and operate those facilities. Fast-forward 50 years, and narrowly intelligent machines could be tasked with the entire process, end-to-end, including their own manufacture and the (costless) manufacture of their supply of energy (most likely solar) and raw materials. There is a substantial literature that has begun to explore the post-scarcity implications of narrow AI, machine labor, and other disruptive technologies. Among the environmental implications, CDR geoeingeering is (in my mind) a particularly salient case. The specific example of self-driving cars merely illustrates that the machine labor in question is not 5000 years away, or 500, but - quite obviously - 50 or less. Best, Adam -- Adam Dorr University of California Los Angeles School of Public Affairs Urban Planning PhD Candidate adamd...@ucla.edu <mailto:adamd...@ucla.edu> adamd...@gmail.com <mailto:adamd...@gmail.com> On Tue, Sep 6, 2016 at 5:46 PM, Michael Trachtenberg <mi...@aesop.rutgers.edu <mailto:mi...@aesop.rutgers.edu> > wrote: HI Adam, The majority of physical chemical processes while controlled will not be accelerated greatly beyond known maxima simply by applying computing capabilities. Mike Michael Trachtenberg, PhD Visiting Scientist Department of Chemistry and Chemical Biology Rutgers University New Brunswick, NJ mi...@aesop.rutgers.edu <mailto:mi...@aesop.rutgers.edu> 609-610-6227 <tel:609-610-6227> On Sep 6, 2016, at 7:08 PM, Adam Dorr <adamd...@ucla.edu <mailto:adamd...@ucla.edu> > wrote: As I explain in detail in the papers I attached and in my other recent work, there are two problems with this reasoning. The first hinges is how we define prudence. Ignoring a possibility until evidence guarantees that the outcome is certain is, I argue, not at all prudent. And the second is that there is already a veritable mountain of evidence that arrival of the specific technologies I described (namely, narrow artificial intelligence and machine labor) is already imminent - to say nothing of the overwhelming confidence we can have that these technologies will have arrived by, say, 2050 or 2075. Self-driving cars are the clearest prominent example, but there are many others. Best, Adam -- Adam Dorr University of California Los Angeles School of Public Affairs Urban Planning PhD Candidate adamd...@ucla.edu <mailto:adamd...@ucla.edu> adamd...@gmail.com <mailto:adamd...@gmail.com> On Tue, Sep 6, 2016 at 2:16 AM, R. T. Pierrehumbert <phys1...@nexus.ox.ac.uk <mailto:phys1...@nexus.ox.ac.uk> > wrote: Yeah, and maybe they’ll get controlled fusion working too. It would be imprudent to bank on such things until there is real evidence that it will happen. On Sep 6, 2016, at 12:57 AM, Adam Dorr <adamd...@ucla.edu <mailto:adamd...@ucla.edu> > wrote: To take just one prominent example, I think that too few folks are giving serious consideration to the explosion in CDR feasibility (and other ecological restoration capacities) that is likely to follow the arrival of widespread narrowly intelligent machine labor (i.e. the AI of the sort that can drive a car, not the general sort that is self-aware and wants to take over the world). Dismissing this as science fiction might have been reasonable 20 years ago. But today, with cars that can drive themselves right over the horizon, I feel very strongly that it is intellectually lazy and socially irresponsible to continue doing so. Other imminent technological changes will also have a profound impact on the feasibility of various CDR approaches. It would be helpful if all who are actively engaged in this arena could take care to avoid some of the more common general errors in reasoning about the future, so that they may think more clearly about the policy, planning, and other implications of technological change. -- You received this message because you are subscribed to the Google Groups "geoengineering" group. To unsubscribe from this group and stop receiving emails from it, send an email to geoengineering+unsubscr...@googlegroups.com <mailto:geoengineering+unsubscr...@googlegroups.com> . To post to this group, send email to geoengineering@googlegroups.com <mailto:geoengineering@googlegroups.com> . Visit this group at https://groups.google.com/group/geoengineering. For more options, visit https://groups.google.com/d/optout. -- You received this message because you are subscribed to the Google Groups "geoengineering" group. To unsubscribe from this group and stop receiving emails from it, send an email to geoengineering+unsubscr...@googlegroups.com <mailto:geoengineering+unsubscr...@googlegroups.com> . To post to this group, send email to geoengineering@googlegroups.com <mailto:geoengineering@googlegroups.com> . Visit this group at https://groups.google.com/group/geoengineering. For more options, visit https://groups.google.com/d/optout. -- You received this message because you are subscribed to the Google Groups "geoengineering" group. To unsubscribe from this group and stop receiving emails from it, send an email to geoengineering+unsubscr...@googlegroups.com <mailto:geoengineering+unsubscr...@googlegroups.com> . To post to this group, send email to geoengineering@googlegroups.com <mailto:geoengineering@googlegroups.com> . Visit this group at https://groups.google.com/group/geoengineering. For more options, visit https://groups.google.com/d/optout. -- You received this message because you are subscribed to the Google Groups "geoengineering" group. To unsubscribe from this group and stop receiving emails from it, send an email to geoengineering+unsubscr...@googlegroups.com. To post to this group, send email to geoengineering@googlegroups.com. Visit this group at https://groups.google.com/group/geoengineering. For more options, visit https://groups.google.com/d/optout.
