Sunday, October 16, 2016

Another Example of a Seneca Cliff: the Demise of Friendster



The results of a Google Trend search for "Friendster", an old social network. It is a nearly perfect "Seneca Shape," where decline is faster than growth.


"Friendster" was a social network that, in many ways, pre-dated Facebook. Friendster collapsed rapidly, starting in around 2009, providing us with an impressive example of a "Seneca Shape", a curve where decline is much faster than growth, or, as Seneca the philosopher said long ago, "ruin is rapid". 

The demise of Friendster has been studied in at least two recent papers. One by Garcia et al, (2013) "Social Resilience of online communities" and another by Yu et al., (2016) titled "System crash as dynamics of complex networks" These papers interpret the collapse in terms of the dynamic evolution of a network, whereas, earlier on, I had proposed a model where this specific shape could be derived from system dynamics.

Basically, there must be more than one way to skin a platypus: both the studies cited are based on collective feedback effects, which is the crucial factor that makes collapses occur. So, network theory is more detailed, system dynamics is more aggregated, but we are describing the same phenomenon, although from different viewpoints.

In both cases, anyway, we find that ruin is rapid. As long as it occurs to an obsolete social service, it is not a big problem. But if it were to occur to something massive and vital for civilization, such as the oil industry, then we could see something like this.... ouch.......









12 comments:

  1. The demise of Friendster was due in no small part to the dominance of Facebook, whose increase in popularity coincided roughly with the former's crash. So though the shape and form was similar to other Seneca style curves the reasons are very much different from say, the impending crash in oil production. I would venture to say that the disuse of obsolete technologies would likely to also resemble Seneca curves..

    ReplyDelete
    Replies
    1. Yes, this is also an interesting point: the models don't "know" about facebook. They only assume a certain tendency for the nodes to leave the network until, at some point, the whole thing goes crashing. So, I think that the Seneca curve can be caused by several different factors, but still maintain the same shape. In the case of the oil industry, it wouldn't be facebook, but the decision of investors to switch from oil to something else, or go bankrupt

      Delete
  2. That would be bad.

    But, those last couple of data points are still trending upward, and we don't know for sure that the traectory of the downslope part of the curve will go as fast as the "classic" curve.

    We also still don't know precisely when the downward slalom will begin in earnest. So it remains a guessing game unless we have a better model for what is going on right now.

    RE

    ReplyDelete
    Replies
    1. It is not that we don't have models, it is that the trend is still uncertain. Anyway, the latest points clearly indicate a flat production. We'll see......

      Delete
  3. As Ugo has drawn it, the area under the curve still represents too big a hit of C02 coming in the next 20 years. When ouch! is exclaimed humanity needs to recognize a bigger ouch is coming if we don't crash oil production even more precipitately.

    ReplyDelete
  4. The context for both social online networks and oil is a complex society. Aren’t supply chains just marvels of connectedness these days? And I understand there is something called ‘Shadow Banking’. I suppose we call it ‘globalisation’. What might cause a Seneca curve for globalised transfers?
    I have been thinking recently about the transition across Europe from the Bronze Age to the Iron Age – I guess it was ‘Seneca-like’. One factor might have been the relative distribution of the resources needed for manufacture. Iron tools were not especially more effective than bronze, I am told. However, most places had some iron ore and fuel not too far away which could have been convenient when trade went down. Suddenly there were defensive ‘Oppida’ or their smaller cousins on all the hills. I look across a plain that previously had monuments and places of pilgrimage for several thousand years (Neolithic to early-Bronze Age) where I now see a ring of Iron Age hill forts.

    ReplyDelete
  5. Oil production is not a network and contrary to social networks, the relative demand does not diminish with shrinking size, but increase.

    ReplyDelete
    Replies
    1. This is a very good point, and therefore, I believe that there is no correlation between how steep the ways up and down are. If the way down is due to a technological breakthrough (for example the introduction of the flat screen was a major change), the way down of the old technology will be very steep, but for technological evolution, the way down of the older technology takes more time (touch screen for mobile phones did not change so much the concept of mobile phone). In the case of petrol, I believe that we will be more on the evolution path, but I agree that arguments go il all directions.

      Delete
    2. Well, to be more precise about my post above,I believe that there are many applications with a high added value where an expensive petrol is ok, for example if you buy a 50 EUR polyester jacket, there is probably not more than 1 or 2 EUR of direct petrol in it. For other applications like heating houses, we already see a major shift in energy source to natural gas, wood, heat pumps, geothermy, solar or electrical cogeneration. So as time goes on, low added value use of petrol is reduced and price can go up, which allows more expensive sources.
      On the other hand, there are two other factors : geology and politics (civil wars...) that can create major disruption in supply, or not give the time to adapt to a new situation.
      I believe that our bigger challenge is that it is very difficult to define what is sustainable and what is not. People and politics tend to forget maintenance and running costs when they decide about investments, but this is a general problem of our fast ROI society.

      Delete
    3. Concerning your list of energy sources to shift to: energy saving is missing. Both in cold and hot climate zones, proper heat insulation plus solar can push heating/cooling fossil fuel needs to zero for a fraction of the total building costs.
      Similar arguments hold for transportation: even with internal combustion engines, new engines like achates opposed piston engines plus downsizing can slash fuel consumption to 1/2, with the same car*km.

      Delete
    4. Dominik
      I agree that 'oil' is not a network, but I suggest the industrial world that can use it for multiple activity (indeed as a 'multiplier') and for economic growth, is very much a complex network; increasingly so in recent decades, witness an expanding WTO and all that.

      Delete
  6. I would not mind seeing another 30 or so examples of the Seneca cliff.

    ReplyDelete

Who

Ugo Bardi is a member of the Club of Rome and the author of "Extracted: how the quest for mineral resources is plundering the Planet" (Chelsea Green 2014)