Accelerating Change

Ever watch old movies? There's a film noir series showing now at a repertory theater in New York, and I've enjoyed seeing several of them. One thing I noticed is how much different is the world we live in today. Watching films from the 1930's or 40's is a good way to get a feel for what life was like then, and how much things have changed. In addition to being good entertainment, those movies also serve as documents of a bygone era.

Not only are the hairstyles and clothing different (fedoras!), but it's interesting to see how much technology has changed our world. In those old movies, there's almost no sign of plastic, for instance. There is no fiberglass or formica, no cell phones or PDA's, and certainly no computers. No transistor radios, or cassette tapes, or flourescent lighting. The cars and trucks are big, clunky gas hogs with no seatbelts, airbags, catalytic converters, or GPS systems. There are no microwave ovens, no frost-free refrigerators, virtually no central air conditioning, and almost no television. I could go on and on, but I won't.

I bring all this up to make the point that technology change is a fact of life in our modern society. In fact, it can be argued convincingly that the rate of change is itself accelerating.

Seeing all this, it's not so hard to imagine a similar level of change in the next decade or two. We might not be able to predict exactly how things will be diiferent, but we can say with confidence that the future -- even the near future of 10 or 15 years from now -- will be strikingly different. And we can commit ourselves to learn as much about this change as possible, and to try to prepare wisely for it.

Mike Treder

Comments

[jim moore]

Not all types of technological change is of equal importance, from the viewpoint of Human Ecology it is changes in Contact Technology that enables / causes major social reorganization. (Human Ecology defines a social system as the spacially bound reoccurring patterns of behavior in a population.) Contact technologies are basically transportation and communication technologies. These technologies are considered to be the most important because they allow the daily patterns of behavior of spacially separated actors to be coordinated.

So, pay attention to how rapidly decreasing communication costs are (can) change your daily patterns of behavior, if you want a deeper insight into the major social changes that are happening.

[Michael Vassar]

I know the arguments, but how can you seriously claim (based on observation, not theory) that the rate of change between today and 1954 is obviously greater than that between 1954 and 1904? I would say that it may be greater, but it's obviously NOT vastly greater. When you use the theory (from Kurzweil, Big Science/Little Science) you have the rate of change doubling every decade or two. This is absurd. I remember the beginning of 1985, and it is absolutely clear that the change between 1985 and today is not greater than the change from the neolithic to 1985. Even if you assume that the rate of change of the rate of change is accelerating you don't dig yourself far out of Kurzweil's hole. You just end up arguing that change since (say) 1970 was greater than change between 1800 and 1970. This is almost as absurd as the first claim. The rate of change in life expectancy and the economic growth rate have been almost perfectly constant since the 1870s. These are the most obvious general markers for change. The absurdities come from using disingenuous markers such as scientific papers published or PhD's granted or patents, e.g. from confusing inputs with outputs. The strange appearent conclusion is that inputs contribute little to outputs, a surprizing conclusion, but one consistant with the arguments from "he who would pay the piper must know the tune" by Robin Hanson.

[Tom Craver]

Change does not seem to accelerate, so much as it is somewhat uneven. In terms of fundamental, society altering technologies, it's hard to beat the impact of fossil fueled engines - largely responsible for shifting us from an agrarian to an urban society.

Automation is slowly shifting us away from the factory and offices and into services. Networked computers are just the latest aspect of that. That's already pretty far along in the West, but cheap labor in developing nations has delayed this trend - back in the 60's we thought robots were going to create a no-work society, but robotics and AI proved more difficult than anticipated. (Let's hope the same isn't true of MM.)

We're facing one fundamental challenge this century that changes the rules that enabled past progress - namely fossil fuel depletion. That's going to take a lot of adjustment, whatever path we take. In retrospect, it will probably appear that the most fundamental changes of this century will have been driven by that, regardless of what technologies are developed to implement those changes (including nanotech).

[jim moore]

Michael,
I think that it is important to differentiate between simple technological growth and technologically induced (enabled) social change. A good example of simple technological growth is: the rapidly increasing number of chemicals that we know how to synthesize. (I think this is the type of change Kurzweil is talking about.) Tom Craver’s example of the internal combustion engine leading to (allowing for) rapid urbanization is a good example of technology inducing (enabling) social change.

Tom,
Hate to nit pick but we are not running out of fossil fuels (there is plenty of coal) we are running out of cheep oil.
And I would argue that it is rapidly falling cost of communications and computing that will drive (allow)rapid social change this century.

[Tom Craver]

Jim - I've made it clear in previous posts that I understand that we're not out or nearly out of oil, but it's a distinction that makes little or no difference, especially considered over the interval of the coming century. We're going to need to get by on less...and less...and less - and that will be a major driver of events in this century.

[Michael Vassar]

A Century!?!
You're willing to make predictions a century out?
Good luck, you'll need it.

[Chris Phoenix, CRN]

Jim, that Human Ecology concept is very interesting. I have a couple of questions:

You said, "Human Ecology defines a social system as the spacially bound reoccurring patterns of behavior in a population." Now, suppose everyone spent all their time online. Would that diminish the importance of the "spacially bound" qualifier? (BTW, is "spacially" a technical term, or a typo for "spatially"?)

Do weapons count as a Contact Technology?

Chris

[Brett Bellmore]

Tom, while oil is running out, given a good technology for CO2 sequestration, we could have a fossil fuel based economy without any trouble through the end of the century, based on coal, tar sands, and (speculatively) methane hydrates. The extent to which we'd have to tear up the countryside to get at them is regrettable, though.

[Mr. Farlops]

The bit of handwaving in the "acceleration of accelerating change" argument that's always interested me is intelligence amplification.

The thought goes--first language and tool use emerge, then writing, then libraries, then the printing press, then telecommunications (telegraph, radio, television, satellites, etc.), then the computer and computer networks emerge. All these, the argument goes, amplify human mental power.

The thinking goes, we now have many ways to suppliment long term memory, to make correlations, to share knowledge and discoveries, to analyze designs, to speed data interpretation. This all speeds the rate of scientific and technical advance, which only feeds back into the loop.

But there has to be damping happening here somewhere. I'd argue that that while it may be true that we can analyze data much faster than we used to at the same time our improved instruments give us still more data to analyze. CAD/CAM has sped engineering but at the same time, engineers have to analyse a lot more failure modes of machine of growing complexity.

Isn't this like the truism that advances in hardware get lost to software bloat? Sure, we advance and we process the work faster but at the same time the amount of work grows so much of the speed is lost.

And how much of human mental function can we bottle up and automate? If it can be stated algorithmically, it can be mechanized. But there are still human mental functions that we don't fully understand yet. For example, can we harness, mechanize and accentuate creative leaps and insights?

I suppose that's really an AI question. Personally I believe the the strong formulation of AI is possible but there is still much neurological research to be done.

There drugs that enhance memory, focus and alertness and these seem to be improving faster of late. Recently there was a study of monkeys where their concentration and "flow" was enhanced. How soon before startups start lacing the coffee of programmers with this stuff? How soon before they start banning doping students before exams?

This brings in the political angle. What if we discover tomorrow that company in South Korea is doping its workers brains to increase its collective intelligence? How Congress, the Pentagon, unions, human rights advocates react to that?

[Brett Bellmore]

As a regular user of CAD, I can identify one of the damping factors: The software I'm using today does some fancy things, but it's enormously less computationally efficient than what I was using a decade ago. I'm using a computer whose clock speed is 300 times faster than the one I started out with, that's got 50 times more RAM than the hard drive capacity of my first workstation, and I assure you, the software doesn't run 300 times faster.

Outside of some applications such as FEA, where computational efficiency is highly valued, most of the increases in computer power have been soaked up by changes in programing styles.

[Mike Deering]

Mr. Farlops, cognitive scientists and AI programmers have many theories of how to automate more mental functions that have not been tested tey simply becuase the computing power necessary is too expensive or unavailable at any cost. When cheap nanocomputers become available we can expect great leaps in AI to follow shortly.

[Chris Phoenix, CRN]

Why do people write whole applications in interpreted scripting languages instead of C++? For the same reason they used to write in C++ instead of assembly language. It's easier and lets the programmer do more faster. PHP has one data structure that's an array, list, hash table, and dictionary.

Don't sell short the "fancy things" your software does. Many of them are marketing checkbox feature creep. But some of them are real innovations.

It remains to be seen whether the flood of programs (and programmers) will actually advance computer usage. Likewise, I'm not sure that it's ten times better to be able to synthesize ten million rather than one million chemicals. But my point is that the computer power is being used, not wasted. It's retroactively saving time for the programmers.

Chris

[Brett Bellmore]

I'm sure it is saving time for programmers; Not consulting the customers about what features we want retained or eliminated saves them time, too. I was just pointing out that increased hardware performance doesn't translate to increased software performance, not one to one at any rate. Most of the computer power on my desk is in effect being used by a programing team at Autodesk, rather than ME.

[jim moore]

Chris,
If you are interested in the Human Ecology approach to sociology, I would recommend a short but dense book written by Amos Hawley called Human Ecology: A Theoretical Essay. I intend to reread it this Christmas break if I can find my copy.

The military extensively uses transportation and communication technologies but in general I would not consider weapons as contact technology. (although I suppose that you could consider a gun as a type of transportation technology for delivering lead)

OPPS that was a typo, it is spatially not "spacially".

[Tom Craver]

Brett:

I'll stick by my prediction. We've got maybe 50 years of oil let at current consumption rates. At a minimum oil, prices will have to go high enough to drive substitutions - e.g. coal liquifaction. The economic adjustments of much higher oil prices will be large.

I do expect MM and nanotech will eventually drive up energy efficiency - but for most of the world, that will translate into improved quality of life rather than any reduction in fossil fuel demand. Decreased transportation of goods will likely be more than made up for by demand for personal leisure travel.

[Marc_Geddes]

Michael,

You said that 'economic growth rates have been constant since the 1870', This is a highly misleading statement.

It's true that in *percentage* terms the economic growth rate has been constant - around 3%.

But a fixed *percentage* rate is actually an exponential growth rate in *absolute* terms.

Think about it. If in absolute terms the quantity of something is 100 units and it grows by 10% in a year, then it the end of that year there will be 110 units - an increase of 10. But grow that by another 10% the next year, and at the end of that year there will be 121 units - an increase of *11* units, not 10. And so on.

So in fact, the world economy has been growing *exponentially* in absolute terms, even though described in percentage terms the rate has been a stable 3%.

[Brett Bellmore]

Tom, if you look at historical oil prices, adjusted for inflation, domestic

http://www.fintrend.com/ftf/images/charts/Oil_inflation_LG.htm

Or international,

http://bigpicture.typepad.com/comments/2004/08/oil_inflation_a.html

you can see that while current prices are high, they are not unprecidentedly high. And, frankly, if the current high is sustained, rather than dropping back down to historical averages within a couple years, I expect we'll start to see some movement towards substitution. I think it's already high enough to make coal liquification financially feasible, if investors thought the price was going to stick. That's the real obstacle, that everyone who might invest in the equipment to do that expects the price to drop again too soon to recoup the expense.

[Tom Craver]

Brett - When I say that process of depletion of fossil fuels will likely be seen as the major driver of changes in this century, perhaps you think I'm saying the sky is falling. No, I'm just saying that, as the onset of coal and then cheap oil drove massive changes in our economy and ways of life, so too will the transition away from oil, and eventually away from coal.

A change to coal liquefaction would be an example of the sort of massive changes I'm talking about - new processing plants going up, coal mines expanding, oil imports declining, eventually less need to fear terrorism as the US feels it has less to gain from being in the middle-east.

Current prices may be sufficient to make coal liquefaction a serious alternative - but I suspect oil prices will have to hit at least $65 (~2x break-even for liquefaction) to stimulate the massive investments required. That's several years off, and in the meantime we'll be seeing the impact of high oil prices - possibly a reversal of this rather weak recovery once investors decide oil prices are "real".

[Michael Vassar]

I don't get it Tom. We're talking about indefinite life extension, etc, and you are calling the construction of an infrastructure for coal liquification a Massive change? Huh?

Mark Geddes: Economic aggregate output has historically increased at a linear exponential rate, but that is at best a linear rate of change, and more realistically isn't even linear. The difference between $1/day and $2/day is much more substantial than the difference between $100/day and $200/day, even though you can call the latter 100x "bigger". For most people there doesn't seem to be any personal utility difference between $10,000/day and $20,000 per day. At this level there remains a status benefit to being twice as rich, but a stock boom that makes you and your peers twice as rich is a wash. In general, with respect to money, the relevant unit, utility, shows sub-linear growth, while an irrelevant unit, dollars, grows exponentially. The diminishing utility of wealth is well verified by casual introspection but is surprisingly extreme when assessed from international polls of subjective well-being. That wealth diminish in value at least exponentially is fundamental to economics. Without this fact no-one would ever spend any money. Instead they would always invest it for returns of exponentially growing value.
Generally, the shape of a growth curve is determined by the unit one chooses to measure. The challenge, to someone promoting an agenda, such as Kurzweil or John Smart, is to find the unit that produces exponential or increasing asymptotic results. The challenge for someone trying to predict and optimize the future is to find the most relevant unit. Here are some other examples. Computing power, the classical example of exponential growth, becomes an asymptotically fading time savings when evaluated in terms of time to perform any given calculation. It grows massively hyper-exponentially if evaluated in terms of the set of potential calculations that can be completed within some time limit, but appears to grow about linearly if evaluated in terms of the set of known useful programs that can be run at practical speeds, if useful is weighted by degree of utility. Engine horsepower has grown geometrically with time, but a more relevant unit, maximum speed, has grown linearly. A more relevant unit still, commute time, has been constant, but the most relevant unit, commute distance, has grown sub-linearly.
You can look at the archives from Phillip Goetz's speach at Transvision 2004 for more on this topic. I have decided to post this paper to wise-nano.

[Tom Craver]

Michael: You're right, there may be some changes arising from MM that are more significant in personal terms - though I see energy as the primary limiting factor on MM, so anything that forces changes to how we get energy and how much energy we have will shape what we can and will do with MM.

As to indefinitely extended lifespan - I guess I don't consider that so huge a change on a personal level. Yeah, that sounds strange - what could be bigger? But looked at another way, it's actually a reduction of change at the personal level - and if it's going to happen in time to save you, you're already living your indefinitely extended lifespan. But indirect effects could be large.

[Marc_Geddes]

Interesting points Michael. It seems that someone gets for rate of change will indeed depend on what units they are using.

The fact of the matter is though, real wealth has been increasing exponentially since the 1870's. A 3% world growth rate translates to the world economy doubling within every 30 year period. So in the period between 1974-2004, there were more goods and services created in real terms than the whole history of the world prior to 1974.

And is it really true that utility has to decline with increasing wealth? It does seem that for most people that tends to be true, but I would put that down to the fact that most people are leading rather unimaginative lives, not because there is in fact any real decline in the utility of wealth as quantity grows.

Let me put it this way: If you or me were or any of the transhumanists here were getting richer and richer, would the utility of that wealth be decreasing to us? I say: No way. I can think of a huge number of worthy transhumanist projects and charities that need our support, and the more money we had to support them the better.

To name a few: Methuselah Mouse, Singularity Institute, Immortality Institute, World Transhumanist Association, CRN ... and so on. For us I do not believe that there would be any decline in utility at all with increasing wealth, not unless it was getting to the point where we were raking in billions of years per year. And even then the decline in utility would be caused by our lack of imagination, not because all the extra money would be any less valuable.

[Michael Vassar]

Tom: You know, I hope, that MM provides for its own energy and energy storage needs. Solar power becomes abundantly cheap, and what energy limitation remains ends up being due to Earth's limited heat dissipation capacity.
Marc: Chris or I could fund projects well into the billions, but only because others are not doing so. Even then, there would be diminishing returns. The second $100,000,000 donated to transhumanist organizations would not nearly equal the first in significance. Also, let me note that it is Power, not personal utility from spending, that effects our ability to promote transhumanist causes. If everyone was 1000* wealthier we would have much less than 1000* our current ability to promote change.
Actually though, I needn't argue. As I said above, exponentially diminishing marginal utility is fundamental to economics. Without it, you would get infinite saving behavior.

[Tom Craver]

Michael: You are aware, I hope, that solar power is a very diffuse and often unreliable resource, that must be converted to another form to be stored? Fossil fuels, by comparison, are quite energy dense, and already in a storable form. So long as fossil fuels are affordable, they will be more desirable than solar power. And MM will not just make solar power cheaper - it'll make fossil fuel extraction cheaper as well.

Nanotech will probably give a short term boost to fossil fuel production - e.g. easing conversion of coal to a clean liquid fuel form. But at the same time, demand for energy will be soaring, as people all over the world increase their standard of living - needing only energy to power their nanofactories, and then having made themselves cars, wanting fuel for them. Even with greater efficiencies, I would not be surprised if global energy consumption soars by a factor of 10x - with a lot of that coming from fossil fuels so long as they are affordable.

Sure, people will shingle their houses with solar if that's easy and fossil-fuel derived energy from the pump or the grid or the farm is getting expensive - but for the vast majority, that won't be enough. Most people will probably take the "safe" route of staying where they are and buying fossil fuel energy so long as they can afford to do so. So again, it'll be when fossil fuels get too expensive that radical changes will likely occur.

[Michael Vassar]

Tom Craver:
There are many ways in which solar energy can be stored conveniently, given MNT. Nanotech flywheels, compressed nanotube springs, compressed gas in diamondoid tanks, hydrogen (with cheaply mass-produced fuel cells) etc. Storing solar energy is no problem, so neither is it's erratic nature, nor is there a need for oil for transportation.
My impression is that you are not thinking in terms of molecular manufacturing here, as in that context your statements are difficult to interpret. 10 times current global energy consumption would still be insignificant to harvest with MNT. Current global energy consumption is about 1.5 * 10^13 watts. Multiplying that by 10 and you will still be using less than the incident solar energy on Utah. Nanotech solar cells can be at least 50% efficient, and can store energy for indefinite time periods with better than 90% efficiency. How do fossile fuels compete with that as an energy source? They don't. Now they may be a useful carbon source. Bulk chemistry will still be needed for efficient production of nanotech feedstocks, but they are abundant for that purpose.

[Tom Craver]

Michael -
I don't understand why you think land area is relevant to the competitiveness of fossil fuels, but your figures appear to be a bit off. Current yearly global consumption is about 4.1e17 BTU, or about 1.2e17 W-hr. ( http://www.eia.doe.gov/emeu/aer/pdf/pages/sec11_7.pdf )

Assume 8000 W-hr per day insolation - higher than you'll get anywhere in the US ( http://www.windsun.com/Solar_Basics/Solar_maps.htm ), and about 2x higher than US average. At 50% efficiency, 4000 W-hr per square meter per day, or 1.4e6 W-hr/yr. You'd need 92600 sq-km, or 300x300km for current consumption or 926000 sq-km = 960x960 km for 10x consumption - bigger than every state except Alaska.