Saturday, 1 October 2011

The Great Stagnation(?) by Tyler Cowen:

This mini-book appeals to me because it talks about economic progress/stagnation/crisis as primarily attributable to innovations. However, his surprisingly compelling idea, to account for our global economic finance bubble, seems to clash directly with Kurzweil's portrayal of miraculously smooth technological advance. (Although Ray does talk about an overall exponential composed from S-curves through successive paradigms.)

I came upon this (£2 Kindle app special) book via a link to this blog post considering stagnation vs relocation (of the world's economic centre to China).

* The Book's Central Thesis in A Nutshell Painted by Me:

The natural course of technological innovation, following the tree of scientific discovery, yielded many 'low hanging fruit' from 1870-1970 that greatly benefited the whole of western society. Transport, communications, home conveniences, mass production, free fertile land for US settlers and perhaps cheap fossil fuels. These rapidly raised the standard of living for just about everyone (in America), doubling it ~ every 25 years, also creating universal employment.

However, in the the last 40 years there have been no innovations with the same level of wide ranging public utility. We, in the developed world reached a plateau of of technological productivity gains in the 1970s. Advances since then have been *marginal improvements* that have mostly benefited private consumers (to an extent dependant on wealth). The lack true economic growth was hidden by the uncertain value of growing government expenditure. [Perhaps outright "Pollyanna Creep" too (my thought, or rather Douglas Rushcoff's from "Life Inc.").] Markets and individuals, caught up in the endemic false expectation of continuously strong growth (from the previous period of reaping 'low hanging fruit') all simultaneously overstretched to the point of buckling.

The Internet is a sole exception of innovation, in that it has had a great qualitative benefit to peoples lives, for that part of society sufficiently educated to enjoy it's many wonders. However, it has not yet raised standards of living as ubiquitously as did electricity in homes. Furthermore, it contributes very scantly to employment and GDP. Computers do almost all the hard work, so massive companies (Google, Facebook) have few employees, hence (in part) the 'jobless recoveries' of the last 2 decades. Also, it's very difficult (or just plain unnecessary), to monetise most internet content. Time well spent on-line, for the individual, may well decrease spending elsewhere, hence reduce GDP and government tax income. The net is just too damn efficient!


* More Details:

We've been subject to a terrible case of diminishing returns in general: real education costs (per pupil) have doubled despite no quantifiable increase in academic ability and the number of school graduates has actually decreased since the 1960s. Health spending has been ballooning, almost exclusively to benefit the elderly, being near impossible to quantify the true value of any care because it is not subject to anything like free market valuation, and by the ambiguous nature of many treatments themselves.

In general, government size and expenditure has continued to increase despite it being unsustainable to do so without the massive productivity boons of the century previous to the moon landings. The contribution to GDP (i.e. our main indicator of prosperity) by government is measured by raw *cost*, instead of by end *value* (like goods, e.g. apples). Therefore, increased government spending increasingly obscures evaluation of our true prosperity. Furthermore, because each successive dollar spent by central government is less effective (on average) than those before (as size of government increases towards 100% or GDP), larger government causes more greatly exaggerated perception of growth (and national well being).

This ongoing financial crisis was a result of: "We thought we were richer than we were". We were fooled by lying growth figures, and complacent from a string of minor financial crises (from "savings and loans" and "Black Monday" of 1980s, up to the "dot.com" of 2001) that never really caused much trouble long term. Politicians had happily backed the poorest in society gathering unpayable debts against overinflated house values, but sub-prime loans were merely the canary in the mine heralding the start of *everything* finally coming to a crunch. From museum owners making overly ambitious expansions, to the banks themselves who had massively overstretched themselves (having leveraged up from 12:1 to 30:1), etc. What's more, the cheap/free pleasures of the internet just accelerated the downturn with people more happily able to avoid some personal spending.


* My Meandering Thoughts (Discussion):

+ What about the advent of computerised logistics in the 80s that brought us the modern supermarket? This provided many significant efficiencies and price-performance improvements for consumers. Though it probably produced a net loss of jobs through the same efficiency savings (putting many more smaller, more specialised, shops out of business), so part of the later 'jobless' technological boons.

[First graph from book.]
Perhaps there is no discrete cut-off of useful innovations circa 1970, more of a gradual sliding towards developments that favour lower employment, higher wealth concentration. Supermarket logistics would kick in around tipping point of this transition. Could it be that GDP figures are roughly right, but the richest have been getting richer by stealing the majority's growth in standard of living? (Inadvertently stifling overall productivity.)

Cowen does state:
"If one sentence were to sum up the mechanism driving the Great Stagnation, it is this: Recent and current innovation is more geared to private goods than to public goods. That simple observation ties together the three major macroeconomic events of our time: growing income inequality, stagnant median income, and, as we will see in chapter five, the financial crisis."


Although Cowen never gives any specific examples of possible future innovations that will benefit all society, I got the impression he expected there to be some in the future, that we are just on the flattened middle of an s-curve or progress. That either the discoveries to be made will happen to be fundamentally egalitarian in nature, or society will start gearing their application thus.

But what if the slide towards greater wealth inequality can only  increase with forthcoming technologies? After all, each successive level of automation raises the threshold of intelligence/capability required to enter the workforce; what happens to the (less academic) majority of the workforce when all shelves are stacked automatically and check-outs scan items while still in a trolley? What about when every home has a nano-replicator that can print *anything* and 'earning' is limited to those who own the 'interlectual propperty' (ok maybe that's looking way too far ahead for this discussion). My point is there is an unavoidably expanding segment of society that is incapable of submitting to employment, even if there were job openings for them.

Our Marxist economic paradigm draws to a close: being broken by digital technology. The internet has no static factories ripe for taxation: it's 'factories' often don't even *sell* anything. And this digital 'economy' will continue to grow, well beyond virtual shop fronts for distance selling. I wonder if alternative currencies, perhaps even ones based on reputation, or other such abstract measures, can help fill the gap here.

Ultimately, I think,  our technology is going to stop needing homo-sapiens altogether. With that in mind, it's never been a better time to start figuring out what to do with our growing mass of economically disenfrancised individuals, because one day that demographic could encompass all of us.


+ Did the mechanised & chemical horrors of WW1 and the nightmare of atomic destruction that ended WW2 are cited (elsewhere) as badly damaging the public reputation of science in the developed world. Could this have stifled a continued rise in the popularity of scientific/engineering careers, derailing the natural course that would have yielded sufficient advances to sustain (3%) continued growth throughout the last 40 years too??? Or did we just hit an unavoidably slow patch in the universe's tech-tree?? Or *is* the economy just moving East afterall???: Paul Krugman - "The New Economic Geography".


+ The only solution (to Stagnation) Cowen proffers is to "Raise the social status of scientists" (in the west). More in line with how the technical minded are revered in china/India.

I have been in *total* agreement with this for some time, and there are hints of a turn in this tide, like the popularity of "The Bing Bang Theory" TV series, for example (maybe). But then UK government (for one) is *cutting* it's science funding, always protecting short term commercialisation ahead of longer research too. And I'm dubious of commercial R&D, with pharmaceutical giants (for example) accused of shortsightedness due to conflicts of interest.

To be honest, I was already resigned to the west ceding intellectual dominance to China (et. al) straight after economic. I don't see this scenario as something to be avidly avoided. Even 'Chinese democracy' has certain benefits over that of the USA, and it should be forced to evolve still kinder manners for individuals as the wealth of their populace grows. I think we would be alright for US/Europe/Japan to take the co-pilot's seat in the long run. The greatest worry, in my view, is the unpredictability of an economically humbled USA, with it's legacy of global military might.


+ Jonathan Huebner's gaussian curved graph of innovation per person (by global population)... I made a point of *not* flying to the defence of accepted Kurzweilian 'wisdom' here, instead considering this figure with an open mind.
[Second graph from book]
It may indicate that good inventions are becoming rarer; fewer per global citizen. This could mean that either there is a decline in interest in participating in science/technology (which seems, to me, to have been a trend in UK culture during my life time). Alternatively, good inventions are rarer or more difficult to create now (which fits with an observation that cutting edge physics has moved from individual philosophers of natural science to massive team undertakings like the LHC, etc).

The dipping peak of this graph does correspond to global population growth, so one could claim that the *total* number of innovations may still be growing rapidly, it's just not keeping pace with the overall population increase. However, Cowen points out that (1965-1989) the number of *researchers* in the US, West Germany, Japan doubled, tripled and quadrupled (in turn) while number of pattens held roughly steady. So the later seems to be dominant: innovations require successively more work.

This doesn't really surprise me; I know first hand that the more cutting edge solid state physics (so useful for all materials science and computing advances) is increasingly abstract, with details beyond my mental capabilities. And with increasing complexity comes larger idea spaces to explorer (with more dead ends). But I suppose I would assumed that increased computing utilisation would overwhelm such obstacles.


- Perhaps the west is paying the price now (with recession) of getting too far ahead of the rest. We need the collective might of the majority of the world's population to pass the last few levels of the technological invention game. So when Britain was Empiring all over the place, enslaving India and suppressing the Chinese with opium it was like 'we' thought the game was Mario Carts, when really it was Portal 2.

Would 'the great stagnation' we are in now be the end state for our civilisation, if China and India could not eventually catch up and save our collective bacon (see point 1, below)? I would like to think that the West alone would *eventually* figure out all the steps to technological immortality alone, but what if it is not like walking down a long straight road of progress?; we should think of societal networks of increasing size and complexity, so maybe without sufficient population to build a sufficiently big network, humanity as we know it will fail to reach it's ultimate goal.

After all, technology companies have had to agglomerate ever bigger to be able to manufacture the increasingly complex new devices upon which society relies: only 2 major carrier jet manufacturers, only 2 major desktop CPU manufacturers, each requiring ever bigger markets to fund their ever increasing assembly plant sizes. Matt Riddly, in this talk I often reference (from 12min), gives examples of cultures that have become isolated from the rest of the world and subsequently lapsed back to using less advanced tools, simply because they no longer had the population to sustain all the necessary trades to make artefact's of contemporary complexity. Perhaps digital technology changes the rules here, stopping us from forgetting the latest technologies or something more subtle.

Ultimately, what stops me worrying about Huebner's graph, even if it were exactly right and the selection of innovations isn't flawed as Kurzweil claims (in New Scientist back in 2005) or misleading in other ways, is essentially this: human beings may reach an ultimate population limit on earth, but technological Humanity (A.I.s), once created, should continue to enjoy an exponential increase in computational resources available for their existence as technology continues to miniaturise and spread indefinitely. After artificial minds swell our population, innovations per *person* is going to get seriously outdated fast.


+ This raises some interesting questions about reaching Singularity, along the lines of those raised in Vernor Vinge's "Marooned in Realtime":
1) What if asia (the land mass and peoples) never existed, would the world still be able to reach technological Singularity with, say, only 50% of the population carrying capacity?
2) Would Huebner's rates of innovation per person hold in a world with a lower population carrying capacity, that world would  experience a more gradual exponential curve towards Singularity? Or would it hit a brick wall and stop; is there am absolute threshold population for achieving singularity?
3) Is there a global rate of advance that is too slow to reach Singularity?
4) What are the failure modes (if they exist) when progress is too slow? (Annihilation? Stagnation and regress? External threats: e.g. asteroids, volcanoes.)
5) Can progress be too *fast* (in a world with greater population) causing failure through dangerous instability?
6) Could (something like) humans have evolved on a planet too small for them to grow to the threshold population, or would other factors (such as strength of gravity too low for atmosphere) prevent this in reality? Are there worlds out there stuck permanently in an paradigm of Roman Empire level civilisations because they have total habitable land masses smaller that that of Europe. Or does that species eventually become extinct to be replaced by a smaller one with more efficient brains and so find the limited biomass sufficient? (there must be a hard limit to biology somewhere)

* A Kind of Conclusion

This book felt like another missing piece of my understanding of reality was falling into place; a joining of my recent recurring interest in economics with a contrasting view on technological progress. Having said this, neither Cowen's vague optimism, nor Huebner's graphs have shaken me from expecting Kurzweil's predictions to materialise, broadly speaking. However, these meditations do add a hint that the flavour of future realities my taste a little sour to many. Also, this book makes a good effort to explain the financial crisis where I've not noticed any convincing explanations from Ray. A fellow Singularity Summit speaker and sponsor is Peter Thiel, who's also referenced by Cowen for blazing the way for his book. I've already discussed Thiel's thoughts on the lack of innovation in the middle of this blog post about a possible green energy bubble (with Summit talk video included). Though it is odd listening to financial/technological advice from a guy who captained a $7.2Bn hedge fund into a 90% loss...

No comments :

Post a Comment