Friday, February 18, 2011

jobs policy and what it says about human intelligence

In January, I posted the following to a discussion list on intelligent systems:


A few weeks ago, Lotfi posted a message arguing that excess obsession with money has become
a real problem in US society, and -- ironically -- may help explain why China is growing faster than the US.

This is not a small matter or a simple one.

In general I agree -- but at this moment I am worried that lack of clarity and intelligence in HOW
people think about money may be the greatest immediate threat here. Of course, these are all
just expressions of personal feelings, and not an official position of NSF.

It looks to me as if major fateful decision have been made in the past few months and months to come.
As we worry about economic growth and keeping up with China, what do we do? Reallocate
a huge amount of limited funds away from K-12 schools and universities to the purchase of yachts.
More concretely -- the tax decisions made last month have resulted in a great surge of yacht sales,
quite literally, and some similar purchases, with immediate benefits to employment. But they also
do nothing to alleviate the coming fiscal crunch in states all across the US (a lead headline in the Financial Times
just a few days ago), which already fired teachers and created furloughs in the previous, smaller crunch.
And I have read this morning of plans to cut back all federal agencies to
FY06 levels -- zeroing out the increase from then to now which was enough to keep this division funding as many as 12%
of normal unsolicited proposals in regular panels.

Does anyone believe that yachts are a more powerful factor of production than education and research?
is that the secret of China's economic growth, a secret fleet of yachts we have not heard about?
Of course, there are all kinds of complex systems which led to this outcome, but in the end, a great value
judgment has been made.

Without getting into more detail -- this is something of a challenge to our efforts to understand intelligent systems.
If the brains of mice and humans were evolved to find an optimal strategy of action, based on some kind
of inborn systems of feelings or utility or reinforcement which aims at survival of some kind of reference group,
how to explain this kind of behavior? Some old style Freudians would say -- "You see, Paul, this proves that
your version of the theory is inherently incomplete. There is not only the survival or life instinct. There is also
Thanatos, the death wish, just as fundamental in the human mind."

Am tempted to say more, but we have a word limit here.

My view is... that what's really going on... is basically just a sign of limited intelligence, lack of evolution of human
brains in certain directions. Above all, it points to different kinds of intelligence, as we evaluate humans
in action today. One kind -- more like IQ -- represents something more like HOW MANY neurons,
operating at what level of intelligence, the mouse part of the brain is exhibiting. I would guess that the folks
making these grossly stupid decisions are every bit as good as I am on many measures of intelligence, such as
vocabulary and agility in solving problems in front of them, let alone winning elections and making money,
which are not trivial challenges. Yet there is a kind of sense of reality that they seem to lack.
It reminds me of the time when NASA worked hard to build an airplane without an engine...

The "sense of reality" has something to do with how humans USE words. It probably has wiring aspects
(as in the NRT gate which John Taylor once discussed a lot), but a major part may be a motivational kind of thing,
such as a tendency to cross-check words and reality... the kinds of features which are relatively new
in humans, and hence not fully evolved. (It's not as if mice make farseeing decisions, after all.)
Two types of intelligence -- how big is your mouse brain, versus how human are you really...

FIRST follow-on:

On Jan 26, 2011, at 7:55 PM, Lotfi A. Zadeh wrote:
There are many people, including myself, who feel that making a living in the United States is much more of a problem today than it was in the past. The good old times are a memory. What is the cause?

I said:

This week, I have heard four important questions relevant to what I posted here on "money and intelligence."

(1) Lotfi's question is one of them. Right now, leaders of both political parties understand that the two variables "jobs" and "deficit"
have terrified the American public, and that those two variables are what really drove the last elections. Since the financial collapse of September 2008,
unemployment has been far worse than it was before, going all the way back to the Great Depression. (With apologies to sports fans, I hope we will
call this collapse "Black September," and engrave it deeply in our memories. Many more people died from this event than from the earlier one.)

This being a tricky question, I will first mention the other three:

(2) "What about the President's State of the Union speech? Doesn't he address your points rather perfectly?"

(3) How do those two measures of intelligence you mentioned -- having a bigger, faster mouse brain versus being more human -- connect with
other measures and work in psychology?

(4) How can systems theory and computational intelligence help us do a better job of managing the world economy?

#2 is the easiest for me to talk about. (Please remember that everything I post here is my own personal opinion, not reflecting anyone I work with or for.)
When a friend here asked me this question, I replied: "Certainly. I would agree with 90% of it. That's unusually high for me; the speech was very encouraging
and reassuring, at many levels. Usually in matters of real policy and difficult choices, I get really excited when people score as high as 50%
in laying out what needs to be done next, at the level they are working at. Lately, it tends to be a huge struggle to get as much as 10%,
and people have a right to feel proud if they get as much as that. I just hope that people really follow through on what he said."
(Caveat: I went off to bed somewhere in the Iraq/Afghanistan discussion, so I can't comment on that part one way or another.)

My friend then asked: "So what is the ten percent?" I replied: "It would be obscene to dwell too much on that 10%. In 2009 and 2010,
I remember how many times I wished that Congress could have just voted in Obama's simple, earlier proposal for climate change legislation,
which I would have scored at only 70%. It was so much better than anything in the mess which followed, and so much better than where we are now."
(OK, I oversimplified. The Cantwell CLEAR Act was not so bad. And a few of the missing details could have become real problems.)

The ten percent: the speech was a bit on the optimistic side in describing where we are now. We DO have great new opportunities before us now,
but that's not the essence of our situation. I heard some right-wing commentators say: "Hey, our K-12 education system is already
failing, despite all the money we've been putting into it. Is this really the time to expand government spending for education?"

The essence of the situation -- as those of you in California know quite well -- is that state and local governments all over the country
are now facing a day of reckoning. This has already resulted in the firings of teachers and cutbacks in universities all over the
nation, but the reckoning we face now is much worse than that.

In December 2010, we all knew that the US has limited resources now. We simply can't do everything.
Back when Bush became president, and the US was running a surplus, it seemed reasonable to give
a one-time Christmas present of big tax breaks for people getting over $250,000 per year, to share in the general prosperity --
but things do change when we are fighting for our very survival. How can we say we do not have enough money to avoid firing all those
teachers and cutting back on all those universities, and then turn around and say we have enough to give a vast Christmas present
to folks who then go out and buy yachts? (Yacht sales really did go up markedly after that was passed in December, and now some
people are pumping out hype about "green yachts.") What ever happened to Thomas Jefferson's points about free public education,
a great cornerstone of America's special situation for the past several centuries?

The issue is NOT whether to put MORE money into K-12 or university education. The issue is whether we will have enough
sense of reality to understand that we are on the brink of a vast liquidation of what we already had, and to understand how
important it is that we make the efforts and sacrifices needed to prevent it. Even if that means sacrificing a few yachts we
thought we had as a special Christmas present. Christmas presents won't mean much if the US economy really goes down the tubes.
The President said :"The worst of the recession is over." I say "Maybe, I hope so, but right now we are on track for
a deep double dip." (Not if we really do what he proposed, however, or if we do the kind of tuning which helps more than it hurts or waters down.)
Of course, many economists would have urged the President to encourage optimism and positive thinking, for good reason; yet
perhaps we need more of a sense of reality here to motivate the necessary actions.


That leads naturally to Lotfi's question. GIven the word limit, I may have to postpone getting into details.
In a way, we are addressing a four-variable multicriterion optimization challenge here -- jobs, deficits, growth
and sustainability. Picking school preservation over yachts (just one piece of the situation) is a different way of spending the money,
with no net effect on the deficit in the short term, probably a net improvement in the job situation (because teachers are more stimulating than yachts),
essential to continuing growth, but not a solution to the larger dilemma all by itself. The larger dilemma requires more,
and does require some technical systems thinking, but I have probably exhausted the word limit,
and need to wait until there is more discussion to comment on.

Best of luck,



Probably I should post more on the technical aspects of jobs,
on market design in the electric power, communications and financial sector,
and the link to the way that Europe seems still stuck in a deep "Nash equilibrium"
far inferior to the Pareto optimum it could get to in balancing deficits, jobs, growth and sustainability. But that's for later.

Wednesday, February 16, 2011

Green jobs again

Myth-makers left and right are still making noise about green jobs, pro and con.
So I repost a message sent to the Energy Conversation listserv in 2010,
still up to date now:

The issue of green jobs has certainly attracted a huge amount of exaggerated advocacy analysis,
both from the left and the right. That's why I took some time last year to read through the major research reports, right down to the details of the econometric models, to try to get a decent first order story on what we
really know, posted at:

Quick summary: the devil is in the details. Climate or energy bills could add net jobs or reduce them,
depending on how they are written. For climate bills -- the present uncertainty about
CO2 regulation by the federal government and by states is inhibiting ALL kinds
of investments in the electric power sector. A top guy in IEEE reports that he recently went to a meeting of top utility executives where all they basically said: to invest more (and thus create badly needed jobs,
especially in depressed sectors like construction and steel), what we need more than anything else is
price certainty on carbon. We don't care so much what the price actually is; only that we know it.
The present confusion is the worst of all worlds. Every day we waste in reducing that confusion is a day
of unnecessary weakness, slow growth and human pain all across the US. The American people, like the utility executives, do worry a lot about who in Congress really cares about them or about the country.

But... climate bills COULD hurt the industrial sector of the US, if not offset in ways that could
cause trade and treaty problems.

Congressional briefing on Energy and economic growth

Yesterday (Tuesday Feb 15) I attended a Congressional briefing in Rayburn 2318 --the usual hearing room for the House Science and Technology Committee.

Here is my report on what they said and then my comments...


The big challenge for the US right now is how to manage and change its economy, to try to do
the best we can on four key variables all at once -- jobs, deficits, economic growth and sustainability.

The "energy briefing" today was led by Rob Atkinson, president and founder of the Information Technology and Innovation Foundation, "My cochairs are Warner and Hatch -- nonpartisan." People described Atkinson as the nation's leading expert on long-term economic growth. Like Schumpeter before him, he has shown how most economic growth is due to technology and innovation, which tends to occur in major "waves" which can last 40 years or so. STARTING FROM THE GOAL OF ECONOMIC GROWTH -- he has concluded that new, sustainable energy technology will be the next big wave of economic growth. US economic growth in coming decades will depend above all on whether we can stay ahead of the tough competition in clean energy. He then asserted that Devon Swezey, Project Director for the Breakthrough Institute,
is our top leader today in the "innovation space." He was proud of their joint report,
"Rising Tigers, Sleeping Giant", comparing the US with China, Japan and South Korea in being ready to ride
the next big wave. The key goal for innovation is to make sustainable energy CHEAPER; whoever makes it affordable first will capture a huge market.
He also expressed support for what was said in another briefing last week, Securing America's Future, especially by Blair (recent DNI) and on behalf of Fed Smith (FedEx), stressing the national security goal of independence from oil.

Swezey gave an overview with lots of striking numbers. Success in riding the wave depends on COMBINING strength
across three key areas:
1. R&D: who invents the new product;
2. deployment: who creates markets to buy it and use in volume;
3. manufacturing: who makes it. The US is now 'way behind the others on (2) and (3), but has kept a narrow lead -- under threat --
in the R&D. The success of the others is not due to some kind of cultural magic, but to well-documented public investments.
We spend slightly more in total in energy R&D for now, but the others spend much more as a share of GNP, and China has been raising
the R&D by 20% per year for 20 years; we have recently seen a huge wave of outsourcing of R&D to China. We need new action in three areas:
(1) more R&D spending, especially aimed at technology we can sell to developing economies, the biggest new open markets;
(2) a new explicit manufacturing agenda, integrated with the R&D agenda. Like Sherrod Brown's The Impact Act;
(3) a major new deployment effort, like Bingaman's Clean Energy Deployment (fund) idea. He gave some interesting examples of R&D outsourcing to China:
(1) Applied Materials, the world's leader technology in ability to make solar cells, just set up the world's biggest R&D center on that -- in Xian; (2) GE moving its leading power electronics and control R&D to China.

Atkinson cited NSF data to show that R&D outsourcing is not just in energy, but across the board. "17 is not enough": He said that the President's proposal to increase R&D tax credits to 17% would lift us from being the 17th ranked nation in OECD to being 13th.

William (Bill) Bonvillian, the Director of MIT's DC office, discussed why it will not be easy for the US to stay in the game. We have a good track record of innovation "on new open turf," but not in upgrading a large existing industry like energy. He expressed great optimism about what's going on in the R&D "front end," like ARPA-E, innovation hubs, and some new Chu initiatives. But we have a big problem with the deployment and manufacturing "back end."
Possible remedies: new financing; standards; creating initial markets; development of testbed markets and systems, especially at DOD, which could
be a great place for deploying and using new things from DOE.

Atkinson then mentioned his Clean Energy Working Group, which has frequent meetings, where Arun Majumdar, head of ARPA-E often speaks.

Majumdar began by noting that there is a huge growth going on right now in world population and in income per capita, which the US cannot change.
That creates a huge new market for energy.. if we can make it cheap enough. That's a key goal at ARPA-E.. to "go for a few home runs."
A small team of energetic people, all engineers, many from industry. He gave four examples:

(1) plants are a great source for fuel (biomass), but natural photosynthesis is only 1% efficient; there is a clear opportunity to get much more fuel from the same amount of sunlight through developing new life forms ("black leaves instead of green");
(2) to replace oil by using electric cars, we need to get the cost of batteries down -- not by tweaking today's best batteries (Li-ion, staunchly supported by Li in the politburo of China), but by breakthroughs like rechargeable lithium-air, zinc-air and manganese ion. 50 teams are now competing for funds in this space from ARPA-E, and he has no idea who will win yet...
(3) electric cars could also benefit from futuristic ideas like inductive recharging on the highway... instead of plugging in, receiving power beamed from the road.
(4) cleaning up coal...

Finally, Reed Hundt said: "Yes, double Arun's budget, I'd support that, but I'm working the back end. The US is the only important nation which lacks a
government-sponsored fund (bank) aimed at clean energy technology. We have a detailed plan..."

In Q&A, Layne of Gifford's office asked about how to get more investment in renewable energy, and asked whether smart grid might help.
Hundt was asked whether he has legislative language for his plan; "not yet, but it should be easy enough..."



I agree with the overall story here, far more than I usually do with this kind of event. Their overall sgory fits with other sources.
For example, I have seen the remarkable outsourcing of R&D to China myself. There are many areas where China
is already ahead of the US -- but many where they have blinders, political and other, where they are 'way behind. It is certainly not all
about dollars. In many cases, overly large well-funded vested interests are what keep either nation behind.

With wind -- China's big numbers are due in great part to Europeans paying for it, through their carbon trading systems, which
have already been changed recently in response to scandals.

With R&D -- we certainly ARE NOT capturing so much as half of what we could be doing, even without raising total investment.
DOE has NOT been too basic and adventurous as some asserted. By giving priority either to industry linked
work, or to some very blue sky fundamental science, they have missed the "middle zone." ARPA-E is an important effort to try to fill in, but
it is still missing most of the major opportunities I have seen in this space. Majumdar's first two examples really were VERY exemplary,
but there are other unmet opportunities, and I have seen at least one very big hole in getting to rechargeable lithium-air
this past year. (An unmet opportunity at Excellatron.) The extra cost of plug-in hybrids over conventional cars is due to BATTERIES and to POWER ELECTRONICS about equally; there have been major breakthroughs in power electronics, in recent NSF funding, which also cry out for much
more rapid insertion and follow-on. A new kind of ARPA-E/NSF partnership might be important to fill in in such opportunities.
For low-cost renewable energy -- solar farms using solar thermal look like THE best hope for now, but we don't have anything like
an effective strategic effort to capture our best opportunities (through R&D) to cut that cost in half or more. Majumdar rightly noted that
TODAY's smart grid efforts do little to cut the total costs of renewables or plug-in cars, but there are unmet, hardly funded opportunities
for more advanced type of smart grid efforts that could fill in. (NSF has funded a little of that, but the relevant pot of money has been decreasing
thorough time; it is too small to be a "BIG STAKEHOLDER.")

For deployment -- a lot of what we need for the car part of that is new legislation, which we have discussed in great detail already.
Another possible way to improve "integration" of R&D with deployment might be to have some kind of liaison system at places
like ARPA-E and NSF. In 1990, when electric utilities were permitted to send up to 1% of their revenues to the Electric Power Research Institute
and include it in their rate base, EPRI had a permanent liaison in the NSF building, looking for good opportunities. Perhaps
some larger clean energy investment consortium or the like could do something similar.

The DOD testbed is certainly important. I have had long discussions with people in that space.... There are great opportunities, but also some very problematic barriers we could discuss in more detail later.