Saturday, April 16, 2016

Cancelled Flights, Biodiversity, Infinite Sums, Heart and Soul

This was originally posted June 12, 2009 at 7:48am, by the same name, on another blog platform.

I’m at Ithaca Airport – my flight delayed, but that means some think time. There are times that an airline, and I’m thinking about one in particular, cancels flights and may chalk it up to mechanical problems when the truth is (I’m sure!) that the flight was under-booked. The “canceled” flight may be mine, which affects me directly, or it might be one “upstream”, which affects me indirectly as a flight delay (because they used my aircraft upstream for another flight). What’s odd is that when a flight is cancelled there is no statement about why it’s cancelled, and I don’t see anyone asking why it was canceled, at least not anymore. Sometime after taking the NSF job I actually went to the ticket agent and politely asked why the flight was cancelled and he was visibly surprised that someone asked!!?? After fumbling around, pecking at the computer terminal, the agent answered my query with “mechanical problems.” That’s amazing, because 20 years ago the agent would have announced up front why it was cancelled. Now, I don’t know for sure whether any particular flight is cancelled because of under-booking or perhaps the need for the airplane to serve a more heavily-booked flight which would otherwise have to be delayed, or the like, but thinking of it this way, be it true or not, has helped my serenity, because the canceled flight won’t produce Greenhouse Gas (GHG)! Carbon calculators will often reflect a savings for YOU if you fly less, but the only way that savings happen is if a flight is cancelled, and as people fly increasingly less, flights are eliminated entirely.

I’m now at Philadelphia Airport, my delayed flight having resulted in a missed connection, and now another wait. I’m serene. 

I had expressed concerns previously that biodiversity wasn’t on the map of the to-be-sustained from my experience at other venues, but MOST of the research talks at the Computational Sustainability Conference at Cornell reported on using computational and mathematical formalisms to reason about characterizing, monitoring, and protecting biodiversity !! Examples included the various approaches to addressing White Nose disease in black cave bats or fighting an atrocious disease mangling and killing Tasmanian Devils (which despite their name are awfully cute), estimating current and future distributions of a large variety of species, determining “optimal” policies for acquiring habitat, through purchases of private land, for the Grizzly bear, Canadian geese, Black Ducks, and other species (i.e., so-called ‘reserve design’), using computer vision to identify whale individuals, and GPS tracking to monitor seagull activity. A few hours ago, just before leaving, I listened to a coral biologist describe a differential equation model of disease transmission among Caribbean coral and their symbionts (new word for me) – the biologist is seeing rapid and pervasive evolution in coral (and pathogens!) through host-pathogen interactions, and this is but one example of moving targets in ecological research that makes computational and mathematical tools for analysis so important for handling complexity. Besides the ‘moving-target’ (aka dynamical) nature of ecologies, uncertainty in observations (how many geese were really parked at that lake?), and competing interests (the bear’s, the rancher’s, the tourist’s, the native American’s, the salmon’s, ….) as expressed by multi-objective functions, all make biodiversity questions complex. It’s a safe bet that the computer scientists and mathematicians are working on these questions because they ARE tough, BUT NOT TOO TOUGH, as they are problems that scientist’s can make a start at formally expressing. Nonetheless, this group isn’t satisfied with stopping at their current level of complexity, and a few economics talks are signaling anticipated scale-up to full-blown (human) societal questions – more later.

I know that there are a fair number people that would not put biodiversity on their list of priorities, and frankly, if push came to shove, I don’t know where I would put it, because I’m myopic. There was one session yesterday dedicated to the social science of sustainability – the human factor -- with one sociologist saying that it was almost impossible to change someone’s ‘values’, but you had a better shot at changing their ‘beliefs’ (of facts?) – changing beliefs was a better avenue towards changing actions/behavior than changing values. You can perhaps show how values are connected – maybe, just maybe, there is a transitivity property of values.

Speaking personally, its good to know that there are those investing their intellectual and physical efforts in preserving other species – that some egghead scientist spends her or his life worrying about some slug that lives in only one lake on Earth, for example, may be a way that Providence implements a caring for that species – I’m glad for those people, even if I’m not one of them – I respect them – I want to help them. Perhaps that’s how a caring for a species piggybacks on my caring for a person??

It’s also been interesting to see how concern with sustainability is implemented in mathematical and computer models. Let’s say that we want a species to persist into the future – into the very long-term future. Ecologists told us that one way they represent this long term assumption was that they summed the *discounted* value of having that species around to infinity UNDER DIFFERENT CONDITIONS, and then they (e.g., USGS) would be biased to take actions that would lead to conditions that maximized the estimated sum of values (or rewards). What is meant by “discounted” is that today brings greater reward than tomorrow – for example, MY being above ground today is most important; my being above ground tomorrow is important too, but slightly less important that my being here today – if I’m not here today, then I can’t possibly be here tomorrow. Likewise, it’s important (to me!) that I be here day after tomorrow too, but its value is slightly less than tomorrow’s, which again is slightly less than today’s. Now replace “MY being above ground” with “Species X being EXTANT” and instead of a timescale of days, think of a time scale of years.

You might think that summing over an infinity of positive values, regardless of what they were, would equal infinity regardless of circumstance, thus there would be no way to distinguish the value of different scenarios (except those for which a value term for a year goes to ZERO aka EXTINCT). But so long as the discounting over the years causes future values to asymptote to zero (approaching zero, but never quite reaching it), the infinite sum will equal a finite value (and there are probably other conditions in addition to approaching zero that are required to achieve a definite finite value for an infinite sum). This was all intuitive to me, but still I had reservations with future reward approaching ZERO … could we have a mathematical conceptualization in which having a species around was NOT infinitesimally close to zero? And so on the last day I was affirmed (which is always good), when an economist got up and talked about non-discounted reward valuations that were possible (presumably, not based on infinite sums?)– the distant future did NOT approach zero value -- apparently well known stuff to some!! This was a great illustration of why this conference was such a good idea. 

Discussion following the economist’s talk got into issues of valuing the current generation (YOUR/OUR experience), the previous generation (PARENTS), two prior (Grandparents), the subsequent generation (Kids), and two forward (grandkids), and its interesting to consider whether mathematical models that value a window of five generations (your grandparents through your grandkids) would lead to policies with long-term benefit (because this window of five would slide down the generational stream so that while our generation doesn’t explicitly worry about our grandkids’ grandkids experience, eventually someone will). The whole discussion reminded me of the Great Law of the Iroquois that states, "In our every deliberation, we must consider the impact of our decisions on the next seven generations." (and on which the company “Seventh Generation” takes its name), and on a quote I heard at the OECD conference a few weeks back in Copenhagen: “we don’t inherit planet from our parents, we borrow it from our kids”, imploring us to leave a better world to our children. In talking to one of the other participants, ML, we were also reminded of the social scientist’s talk of the day before about the near impossibility of changing values – but there might be critical points, corresponding to births (and deaths) of those we love where people are particularly open to value shifts – ideally, for the benefit of long-term sustainability of the planet we know and love.

Complexity, Computing, Citizenship

This was originally posted on June 7, 2009 at 9:33 pm by the same name on another blog platform.

Here is a quote from an article ( on the conference I'm attending tomorrow:

"Conservationists will report on using computers to find the best balance of many competing factors in deciding, for example, which tracts of land to purchase to preserve a species, taking into account the budget, cost of parcels, economic impact on nearby communities and the best interests of the species. ... In projects like these, computers offer a way to try out thousands or even millions of possible scenarios, ..."

The scope of the conference is much broader than this one example, but all or most of the projects to be discussed will be looking at complex environmental and societal problems with multiple competing interests -- interdependent interests that have to be (or should be) balanced.

As my plane was touching down in Ithaca -- I mean just as the wheels hit the ground, I was reading this from "Soul of a Citizen" (Paul Loeb), quoting Wendell Berry:

"A bad solution is bad, because it acts destructively upon the larger patterns in which it is contained ... because it is formed in ignorance or disregard of them. A bad solution solves for a single purpose or goal, such as increased production. And it is typical of such solutions that they achieve stupendous increases in production at exorbitant biological and social costs."

and after more acknowledgement of Berry, and Loeb's inferences that "good solutions ... are part of a larger whole", Loeb goes on to say "We can contribute to the well-being of our society, the body politic, by applying a similarly holistic ethic of interdependence, and by listening to those whose voices are too often excluded from public discussion." (p. 131, Soul of a Citizen).

This reading was striking because acknowledging and balancing interdependent factors, concerns, and the like, is largely what this conference on Computational Sustainability is about, and using computer power to converge on and manage complex balancing acts which long-term sustainability will require.

I'm going into this conference, having just finished the conference of policy makers in Copenhagen on Information Technology and the Environment, with several "big" questions. One is what are the different meanings of "sustainability", if in fact anyone has a good definition? Exactly what are people expecting/wanting to be sustained? Of late, I've been disappointed that biodiversity doesn't seem to be something that is an explicit part of discussion -- it seems that as the environment worsens, values that we thought we wanted to be sustained fall away and we are "happy" with increasingly impoverished outcomes. This all fits the addiction model pretty well.

One point that I take from Loeb is that human decision makers ignore or deliberately remain ignorant of factors, sometimes, because they just don't know how to deal with them -- they'll even lie -- blatantly lie -- so that they can avoid the overwhelm caused by complexity. I'm hopeful that computing that enables us to deal with great complexity by processing millions of scenarios will also let us hear and consider all voices, be it the grizzly bears', the ranchers', the tourists', or in another setting, the homeless or the uninsured -- this is, perhaps, a humanistic possibility of computing.

Copenhagen OECD Conference 2009

This was originally posted on May 30, 2009 at 11:31 am on another blog platform as "Copenhagen 2009".

I just got back from Copenhagen, where I participated in the Organization for Economic Cooperation and Development (OECD) conference on Information and Communication Technology (ICT), Climate Change and the Environment. Last year I had been the only US government representative, and that had been a little scary, somewhat disappointing, and very cool – when they called on me for questions and comments they called on “United States” and before opening my mouth I reflected on what I was about to say :-). This year there were representatives from the State and Commerce Departments, and my (big) boss and I from NSF. I’m not revealing any secrets here – last year’s and this year’s are on the Web ( -- more on this later and elsewhere.

I took the redeye from Dulles to Copenhagen late Monday afternoon, arriving Tuesday morning Copenhagen time. I was on Scandinavian Airlines (SAS) and I like it – every seat its own little screen with on demand video and music, a salmon dinner with not-quite-spaetzle pasta and lima beans (!?) to die for, power outlets for laptops, and stuff I’m forgetting. I watched two movies, "The Curious Case of Benjamin Button" (on the way out) and "The Reader" on the way back, both of which I'd balked at when they were in theaters, but I liked them each, particularly Benjamin Button, which was especially moving.

I took the train from Copenhagen airport to downtown, stowed my big backpack in a locker, only keeping my government laptop around my shoulder for safekeeping. Last year I happened upon the National Gallery of Art just before having to return, and wanted to see it especially this trip, but in no rush, and in fact exhausted but with really no option but to stay awake for many more hours, so I walked. 


After a full morning, I took the train down to Helsingor, a beach town where the conference was held, did some brief walking around, talked to the tourist center, then a public bus to my hotel, a spartan place a couple of miles inland. I really wanted to sleep, but there was a conference reception that I thought I should attend for professional reasons, I wanted to stay up as long as I could for a better adjustment to the time zone, and frankly I felt isolated and knew connecting would help, and it did help.

The conference, in a nutshell, was focused on using information and communication technology (ICT) to reduce greenhouse gas (GHG) emissions due to humankind, and most notably what governments should do about promoting and supporting these efforts. There is great optimism that much can be done through virtualization (e.g., telework and conferencing with Web-based computer tools instead of the barbaric practice of flying cross country – yes, I recognize the irony :-), smart embedded systems (e.g., smart cars, smart buildings), which save energy through a large variety of computer-controlled mechanisms, and intelligent decision-making and planning systems that combine climate, economic and other social models. There is a lot of pushback though in implementing so much of this – for example, there are innumerable organizations that require that you keep your computer running 24/7 for reasons of “pushing” software and security upgrades, but my gosh, fixes to this should be 20 years old by now and the stupidity of designing such an energy-inefficient system is stunning -- it offends the engineer in me. I could go on and on with other forms of pushback and lack of awareness, and frankly, among other reactions, our failure to solve even simple, gross technological inefficiencies like the 24/7 computer-on policy has given me a certain sympathy for the disfunction that seems apparent in those trying to solve the planet's truly hard problems. I expect to be writing more on the topic of the conference. Generally, there is a lot that computer scientists and engineers can offer, both in terms of the products that they produce and in the ways that they think – my (big) boss Jeannette Wing has termed the latter “computational thinking”. 

It would have been fun having Pat with me to share the simple excitements and anticipations of a plane ride, train ride, castles, gardens, museums, and I hope that for next year as we are a good travel team, but next year I might be participating over the Web! So many decisions coming up, and even the “easy” ones can give me angst.