Monday, February 13, 2017

Trump's choice of science adviser: something small with huge implications (not just for climate and the soul)

People have told me that the President is now trying to choose between two people -- Happer and Gallanter -- to be the new head of OSTP. Some people would say this is less important and urgent than the other things on his plate right now, BUT IS IT REALLY? (Addendum: it would be so much better if Trump would also consider Lowell Wood, Ed Teller's old right-hand man, as an alternative.
I will add a little on that at the end; Wood would be much better for strengthening America as such than either of the other two... but I do not know either enough now to compare spiritual aspects.)

Some people say it is not such an important job, since it is just a matter of setting up committees on which agencies are represented and come up with public relations pieces to try to justify continued or expanded funding for their various vested interests and iron triangles. That's how it worked under Obama, who got ahead learning to make deals under Harry Reid, a stalwart of the new stakeholder system (aka the swamp).

But the Science adviser position was created for a totally different purpose, in the beginning. In essence, the science adviser was supposed to provide a bridge between serious, concrete technical and scientific truth -- and our very best, deepest and most understanding of that truth -- and the decisions of government, which all too often are not based on reality, in either party. Nowadays the need for that position to work effectively has become all the more urgent, as mass media seem somewhat less coupled to objective reality, and as new risks and opportunities have become larger and larger and ever more imminent.

As it happens, I have deep knowledge of both of the two candidates, due to the way my path was wandered through interesting places at time. I have met Gallanter personally, but I have more comprehensive knowledge of Happer's work. In general: Gallanter may well have aspects which could limit his performance, but he might be able to do the job better than Holdren did; however Happer would be a clear disaster, and a clear invitation for the scientific community to keep its mouth shut or give up on the administration, in several different unfortunate ways all at once.

Details. (A proper OSTP chief should of course be clear about facts versus editorials, data versus theories, and such... as the New York Times was back when I subscribed decades ago...)

1. Happer, climate change and Marshall Institute ============================

In 2009, I started the year as one of two co-equal people working for Senator Specter to handle climate issues, first, and other issues at a lower relative priority. Since Specter was a Republican on the EPW committee of the Senate, this meant we were effectively part of the team of a very friendly guy who worked for Senator Inhofe, who was in actuality the world leader of the movement to cast doubt on global warming. From my past history, you can see that I have had much less hesitation than >99% of scientists to disagree openly with a mainstream scientific viewpoint when my own logic and integrity justified doing so... thinking independently and always questioning. And so I realized that it would be politically convenient if I could really justify climate skepticism, and I had a very strong social obligation to question the mainstream view of climate in depth. As it was my main job to look into climate for the Senator, I did of course read many dozens of papers and try to do justice to the doubter's position.

Of course, EPW held hearings on this issue, and the Inhofe people were clear that Happer was their star witness, the one person who was the real intellectual leader, they felt, of serious reasons to doubt climate change. As a professor at Princeton, he was a real scientist. I heard his testimony, but also read further. I have a photographic memory of sitting in my little stall on the 12th floor of the Hart Senate office building, poring over the words of Happer's testimony. I remember a brief thrill of excitement as I realized that his testimony made logical sense, and that there was hope I could
justify taking that cause further. His central argument was that the greenhouse effect based on CO2 warming the earth's atmosphere has already come close to saturation, so that there would be no further warming beyond what is already in the pipeline. He stated that the climate models use a global absorption rate parameter, not accounting for the way in which CO2 absorbs light in CERTAIN FREQUENCY BANDS, not all light. Of course, the climate denial publications also stressed how the big climate models are all just based on theory, not on actual aggregate changes in a global scale, and they stressed that certain conflicting trends are ignored by the mainstream.

I was excited... BUT LIKE ANY PROPER SCIENTIST, like the kind of adviser the President desperately needs, I didn't just jump onto full endorsement of an initial impression. I knew it was very important, for the sake of truth, to check out and verify some of the claims myself. First, and most fundamental: was it really true that the global climate models are so simplified, and how would they respond?

As it happens, I had previously met a guy named Carl Wunsch, a climate modeller from MIT. Was this just a coincidence? Not entirely. Years before I was invited to give a talk on advanced algorithms for massive data analysis and data mining.

(On my regular web page, www.werbos.com/Mind.htm,  I posted the link and citation:
3. A review of how to calculate and use ordered derivatives when building complex intelligent systems.  I first proved the chain rule for ordered derivatives back in 1974, as part of my Harvard PhD thesis. The concept propagated from there in many directions, sometimes called “backpropagation,” sometimes called the “reverse method or second adjoint method for automatic differentiation,” and sometimes used in “adjoint circuits.” But it is a general and powerful principle, far more powerful than is understood by those who have only encountered second-hand or popularized versions. The review here was published in Martin Bucker, George Corliss, Paul HovlandUwe Naumann & Boyana Norris (eds), Automatic Differentiation: Applications, Theory and Implementations, Springer (LNCS), New York, 2005.)

Wunsch himself also gave a talk there, on how he used backpropagation (aka the second adjoint method) to calibrate and assess his global climate model to the aggregate global data on changes in many variables. I knew from his talk that a lot of the claims you read about from climate deniers are 
grossly false, grossly not vetted or checked... and I also knew that the mainstream "Bible" of the time (IPCC IV) did not ignore the trends which the deniers asserted they ignored. BUT EVEN SO,
the failure of the deniers to live up to the requirements for truth and scientific method DOES NOT BY ITSELF prove that Happer's conclusion was wrong. Guilt by association is also a massive lapse of scientific method, and I did not want to be guilty of that myself ... or to miss a possible opportunity.

So I called Wunsch from our Senate office. Wunsch was a very tough guy, a kind of skeptical curmudgeon by nature, and it wasn't as if I called because we were palsy walsy. I needed all my meager social skills to get clear answers. Wunsch had previously said critical things about his colleagues in the climate modeling business, and had even agreed to appear as one of the stars in the famous anti-Gore "underground" or "samizdat" video attacking Gore which was circulated very widely before then. He was very livid that the folks who made that video had been so dishonest and dishonorable to him, and showed him out of context in a way which made him seem to say things he would disagree with violently. (Perhaps the video makers had lulled him into thinking they too were tough and honest skeptics like him... like my wife? Or is she further to the right in this one? I listen to her too.) His MIT web page was full of his reaction to that video, and it was a huge sore point for him.

But I pushed away from that and from his legitimate feelings about those guys, compounded by the fact that I WAS representing EPW Republicans, and pressed for specifics. Key specific number one:
OF COURSE his model disaggregated different frequency bands. Even his idiot colleagues would not make such a baby mistake. That was a shock: how could we trust a guy who would make such a strong assertion about what's in the climate models, and testify false things even to his supporters and patrons on the Hill, without even checking such a simple basic fact which he had clearly just made up? (No, I wouldn't view the current Flynn flap in the same light, for other reasons... among them being that he is not a science advisor ... but... well, maybe more later on that.)

But then: if separate bands are considered, how could the forecasts come out so different from what Happer would have expected? I hesitate to say a WHOLE lot more right now, because there comes a time when one would want to hold a kind of technical review panel -- IF one had the job, mandate, and money, none of which I have in retirement. But in general, it came down to assumptions about how water and humidity, eta, figure into it. And so I probed the Happer/Homeland Institute work probing the assumptions and implications of the eta variable. Probably the best of the climate deniers would have said that Congressional Hearings must by nature be oversimplified, and (still hoping to support their bottom line), I located a massive detailed manifesto they praised presented at their lead specialty scientific conference, which stressed the impact of eta.

And it was incredibly appalling. Real science requires some awareness of basic arithmetic at a minimum (multivariate calculus and linear algebra even for a competent yeoman level), and that requires understanding you shouldn't say a lot about a variable like eta without defining what it is. Turns out, there are two key concepts, RELATIVE and ABSOLUTE eta, which are drastically different in value and behavior, even though they are conceptually closely related. In a way, this is like the distinction between income in real dollars and income in nominal dollars -- but much, much bigger.  For someone who thinks straight with numbers (numbers like income)... the manifesto was basically a godawful mess, a string of bad puns, ending up with utterly silly and indefensible conclusions. And IPCC IV clearly drew on people who were a lot more clear about such basic issues, which they certainly did not neglect.

By the way, even at the end of 2009, I would have been OK with joining the new subgroup at the Homeland conference which claims that global warming is too small to justify the priority it was receiving. The IPCC IV gave a base case projection that damage not world GNBP would only be about 5% over a century! I was aghast that Reid's staffers rejected Specter's proposal for strong immediate action on oild dependency and the Middle East (see www.werbos.com/oil.htm), a far larger issue, because of his outside agreements to do a Waxman type bill first. Even Woolsey, who advises Trump and has spoken out on oil dependency, simply never told his advisees about the crucial technical realities which would allow much greater progress at lower cost on that basic issue!
But as I continued to ask questions... I learned that we live in a different world from what anyone on EPW knew; see the color figures in www.werbos.com/Atacama.pdf.

And so: if Trump picks Happer, he will get someone who does not probe beneath even the shallowest level of speculation and emotion, even for those issues which Happer pays most attention to. In other areas of science and engineering, it would be hopeless. As head of Marshall Institute, Happer also spearheaded the big reorganization which got rid of their efforts to advance national security... and liquidated new efforts which could have given the US a substantial new lead in what we can put into space.

2. Gallanter ===================================

No great revelations here, but I am amused in part by the fact that I met the guy years ago. I seem to have a very strange kind of luck with such things...  I remember running into a guy named Ed Heath in London, at a kind of cocktail party where I was (for a few minutes) a quiet member of a group of people standing around him, mostly talking about gardening. I remember in 1966 running across Richard Nixon as part of a small group of Young Republicans at a bar in downtown Boston. And Pinkie Bhutto, long before she became Prime Minister  (and was assassinated) in Pakistan. Those are just a few examples....

More seriously, Gallanter is famous for his joint book, Plans and the Structure of Behavior, by Miller, Pribram and Gallanter. That book was famous when it came out as the first important manifesto of the new "cognitive movement,"  a movement to free psychology from the previous "religion" that only Skinner-style rat psychology could be admitted into serious science in psychology. Years later, I worked very closely with Pribram. I participated in many of his workshops, and even arranged funding and organization for a few of them. (Not from NSF, but from the International Neural Network Society, which I led for two years.) Gallanter also came to one of them, and we were all pretty much on the same wavelength. But I did not read his other books, and we did not stay in touch after that.

Would Gallanter immediately know himself the compelling reasons to immediately get rid of the horrible perversions of brain research which Tom Kahlil (working directly under Holdren, Obama's science adviser) was pushing in recent years? The so-called "noninvasive brain stimulation" which uses things like microwaves to directly address reinforcement centers of the brain, letting us turn US soldiers wearing  their helmets to become like helpless junkie/robots, just like what is depicted in Captain America: the Winter Soldier or in the latest Star Wars movie? Sure, I know psychiatrists who complain about crazy people who say the CIA is controlling their brains... but past is not always future... and new technology DOES let bad people do things which were previously just fantasy. Given a more honest science advisor, I hope that James Olds of NSF (who knows the original research by his father on such reinforcement centers) would be willing and able to work with Gallanter in straightening things out, and redirecting new brian initiatives in a way which supports more natural human sanity and potential... despite the very serious and scary "swamp monsters" looking for new markets and new power, watching over folks like Khalil (a guy whose very name reminds me rightly of the kaleds and daleks of ancient episodes of Dr. Who).

===========

As for Flynn... whatever else is going on... some folks have VERY intense double standards
(paying more attention to the tiny mouse of RT versus the goliath of certain Arab investors in Fox News among many other things)   ... who basically want Trump to become defenseless and alone,
in preparation for trying to get to Pence as a more compliant agent of what they want.
However tricky the US-Russia relation, just reverting to acting like mother-in-law to the world and trying to pick a war would really play into the hands of people who very clearly would love for Russia and Iran to bring US and Israel to mutual annihilation, to clear the way for a Third Caliphate
(suzerain for a short period over a US dynasty of some sort, to be assimilated as seamlessly and invisibly as possible).

Do they really want a war?

Hell, just a few weeks ago the head of cybercon urged people to do an immediate demo of our ability to shut down electricity in a major Russian city...

===========

Re Lowell Wood: I do hope people remember Ed Teller. Probably the least guy least soft on Russia, in reality, of all the folks in the Cold War... but scientific enough that he would not go wild. Also,
the experience of figuring out Livermore Lab is so much more real than what the other two have done...


No comments:

Post a Comment