Testing My Growth Model

I have suggested that long run growth can be described as a sequence of exponential growth modes, from primates to foragers to farmers to industry, where mode transitions are similar in their degree of suddenness and growth rate change factors. This model will be tested in the future – it suggests that within a century or so we’ll see a change within five years to a new mode where the economy doubles every month or faster.

But my model can also be tested against the past. Our data on the animal, forager, and early farming eras is pretty poor. My hypothesis suggests that the forager era was one big growth mode similar to the farming or industry eras, with a relatively smooth rate of growth in capacity (even if rare disasters temporarily disrupted the use of that capacity), and that the forager to farming transition has a level of smoothness similar to that of the farming to industry transition.

Contrary to my model, many have suggested there was an important comparable revolution in human behavior around 50,000 years ago. My model predicts that growth accelerated smoothly from around 100,000 years ago to the near full speed farming world of about 5000 years ago, similar to the way growth accelerated from 1600 to 1900.

The latest results seem to support my model:

Back in 2000, a now famous scientific paper called “The Revolution That Wasn’t” argued that the then-conventional wisdom that modern human behavior had erupted in a “creative explosion” about 50,000 years ago in Europe was wrong. Rather, anthropologists Sally McBrearty and Alison Brooks contended that modern behavior, including creativity, has deep and ancient roots, going back some 300,000 years ago in Africa (Science, 15 February 2002, p. 1219).

At a meeting here last month, researchers heard new evidence that human evolution took a gradual, rather than revolutionary, course during two other key junctures in prehistory. A study of ancient stone tools from South Africa concludes that hunters manufactured spears with stone points—a sign of complex behavior—200,000 years earlier than had previously been thought. And new excavations at a 20,000-year-old settlement in Jordan, laden with artifacts typical of much later sites, suggest that the dramatic rise of farming villages in the Near East also had early and deep roots. … Many archaeologists now think that apparent “revolutions” are due to gaps in the record or to behavioral shifts triggered by changing conditions, rather than sudden advances in cognition. What appear to be precociously sophisticated behaviors are really reflections of what prehistoric humans were capable of all along. (more)

GD Star Rating
Tagged as: , , , ,
Trackback URL:
  • V

    Does your model assume a plausible dynamic or is it just statistical extrapolation of time series?

  • Adam M

    I thought the behavioural modernity idea was based on datings of cave paintings, stone sculptures and ritual burial sites, not on the complexity of stone tools. The one site at which they mention finding those kinds of cultural traces dates from ~19000 ya, which actually seems supportive of the BM idea, if anything.

  • V

    And, by the way, what is growth among non-human primates?

    While in some primate species the males give food and grooming services to females in exchange for sex, you could hardly call that an economy with a GDP that can grow. Not even all human societies have an economy with trade.

  • arch1

    1) I wonder to what extent the World Product estimates on which these models are based were really developed independently of the exponential growth functional forms assumed in these models.

    2) That said, if either version of the “sequence of exponential growth modes” is accurate, I wonder whether each transition can be associated with some kind of information revolution. E.g. might the animal->hunting WP growth transition primarily be due to the rise of culture (not just DNA, or individual brains) as a means of retaining / transmitting / developing information? And might the other transitions similarly be linked to significant jumps in intelligent beings’ information handling/generation/management capacity?

    • http://timtyler.org/ Tim Tyler

      Evolutionary theory suggests that relatively sudden changes may be the result of symbiont acquisition – so: memes, plants and animals – perhaps.

  • http://www.gwern.net gwern

    I’m glad you’re testing your theory, but I have to wonder – what data has come in over the past years contradicting your theory?

  • John

    I geniunely call bullshit on that. This is inductive reasoning at its worst. No, actually, it is inductive reasoning using CRAPPY data at its worst. A Popperian nightmare.

    Seriously, this is just one idea less idiotic that the exponential growth graphs that Kurzweil uses (and that have not changed at all for the past 7 years).

  • http://entitledtoanopinion.wordpress.com/ TGGP

    I hope in the future you declare tests before you’ve already seen the data.

  • http://markbahner.typepad.com Mark Bahner

    Why not simply plug in the “population” of “human brain equivalents” into the Solow-Swan growth model?

    Why economic growth will be spectacular

    According to my calculations there, the number of human brain equivalents added by computers will be:

    1 million in 2015 (should not be noticeable)

    1 billion by 2025 (should definitely be noticeable)

    1 trillion by 2033 (should consistently deliver higher growth per year than ever observed before)

    1 quadrillion by 2040 (Singularity territory).

    If adding 1 quadrillion people per year to the Solow-Swan growth model doesn’t generate some incredibly high numbers, there’s something wrong with the model. ;-)

  • consider

    1) Kurzweil did update his graphs to 2009 or so for a presentation, but I cant remember which. It may be the one where he is talking with an Army group.

    2) I wondered the same thing about non human primates, but my guess is that Hanson would argue that there was growth but tiny. I’d say zero growth and then very tiny but…

    3) Intel has argued that computer speed should continue to at least 2029 and this has kept me up worrying some nights….

    4) Does anyone have a edition to Omni that was put out in summer 1989? Some author scared the crap out of me by arguing chips in brains by 2030. It was convincing at the time. That sounds like Kurzweil, but I didn’t see his name in the special Future issue. (a special future Issue seems redundant for Omni.)

  • http://markbahner.typepad.com Mark Bahner


    I confess I’ve always been very suspicious of this “sequence of exponential growth modes” “model.” It seemed to me that “within a century or so” is not much of a prediction. For example, suppose in 2130, economic growth increased dramatically. Robin would be “right,” but most of us would have forgotten (mainly by virtue of being dead).

    But I do see that I’d failed to find the buried lead in another paper of Robin’s, wherein he writes:

    “Without machine intelligence, world product grows at a familiar rate of 4.3% per year, doubling every 16 years, with about 40% of technological progress coming from ordinary computers. With machine intelligence, the (instantaneous) annual growth rate would be 45%, ten times higher, making world product double every 18 months!”

    Hanson on economic growth with machine intelligence (caution: contains math ;-))

    Since he does give a approximate time frame of the next decade or two, I consider that a much more bold prediction. In fact, I don’t think it’s unreasonable, if it comes to pass, to call it the greatest prediction in the history of economics.

    So he’d have that going for him, which is nice…