[HN Gopher] Is growth linear, not exponential?
       ___________________________________________________________________
        
       Is growth linear, not exponential?
        
       Author : jasoncrawford
       Score  : 31 points
       Date   : 2022-04-20 19:18 UTC (3 hours ago)
        
 (HTM) web link (rootsofprogress.org)
 (TXT) w3m dump (rootsofprogress.org)
        
       | robocat wrote:
       | And perhaps there is a new linear section starting with the
       | computer revolution?
        
       | carapace wrote:
       | If you like this do yourself a favor and grab a copy of Vaclav
       | Smil's "Growth - From Microorganisms to Megacities"
       | https://mitpress.mit.edu/books/growth
        
       | georgewsinger wrote:
       | This idea was originally discussed by Alexey Guzey back in 2021:
       | https://guzey.com/economics/bloom/#bloom-et-al-appear-to-not...
       | 
       | Perhaps there's something I'm missing, but it's weird that he
       | isn't getting any credit for it.
        
       | apolloartemis wrote:
       | Can't you model almost anything as a piecewise linear function? I
       | don't know if this claim is saying much of substance.
        
         | melling wrote:
         | I've heard this too. How is modeling as a piecewise linear
         | function best done in practice?
        
         | jasoncrawford wrote:
         | Yes, but if your data is actually exponential, the linear
         | segments are not going to be _better_ approximations than an
         | exponential curve. That 's what's going on here
        
           | ajuc wrote:
           | For any sampled data you'll get guaranteed 100% fit by making
           | it pieceways constant with N fragments where N=number of data
           | points. It says nothing about the function you sampled, it's
           | just a way to cheat by overfitting.
        
           | Fomite wrote:
           | The dangerous bit is that an exponential curve will also be a
           | fairly good fit for a logistic function that's not yet fully
           | observed.
        
             | LeegleechN wrote:
             | Every apparent exponential in the real universe must
             | actually be a logistic or some other bounded curve.
        
           | apolloartemis wrote:
           | True, fair point.
           | 
           | Yeah like the article mentions, they are basically making an
           | analogy to the idea of "punctuated equilibrium" from
           | evolutionary biology. Here's a good exploration of how
           | punctuated equilibrium works, vs the alternative which is
           | called gradualism.
           | 
           | https://gvpress.com/journals/IJBSBT/vol3_no4/3.pdf
        
           | Enginerrrd wrote:
           | I'm not sure that's true in general, nor even frequently. In
           | fact, I'd say it's provably false in general.
           | 
           | The big issue is that you get MANY more curve-fitting
           | parameters to play with if you use a piece-wise linear model
           | vs. an exponential model. (You get to choose HOW MANY breaks
           | to make, what the slope is for each section, and WHERE to
           | make the breaks.)
           | 
           | So... Let's say you created some synthetic data using an
           | underlying exponential plus a normally distributed random
           | number. Obviously, the BEST predictive model is an
           | exponential one. However, for any arbitrary number of
           | observations, I guarantee you there's trivially at least one
           | piece-wise linear model that will have less error than the
           | exponential one. Consider the one that is simply a straight
           | line between EVERY point. Obviously that has zero error
           | compared to the exponential model. Yet, it has very little
           | predictive power compared to the exponential model.
           | 
           | Now, that's not what was done here... but there's actually
           | quite a few parameters in the form of where to make the
           | breaks and how many to make. Doesn't seem like a fair
           | comparison.
        
           | dredmorbius wrote:
           | NB: "Note that both of these charts are on a log scale."
           | 
           | Appears to apply to the two preceding _linear-scale_ charts.
        
           | adamsmith143 wrote:
           | If you zoom in far enough every curve looks linear.
        
             | mathieutd wrote:
             | Not quite!
             | https://en.wikipedia.org/wiki/Weierstrass_function
        
               | apolloartemis wrote:
               | Wow that's super cool, didn't know about this
        
               | foobarian wrote:
               | Heh. Reminds me of how gamers manage to find exploits to
               | cheese speed runs while developers react in dismay.
        
               | AlchemistCamp wrote:
               | Thank you :)
        
       | aidenn0 wrote:
       | Is this new? I remember reading an article a long time ago (maybe
       | 15 years?) discussing how growth in aircraft speed was overall
       | exponential, but linear for each new technology that was
       | introduced.
        
         | mrtnmcc wrote:
         | You can get an exponential growth if the time until the next
         | technology decreases linearly with each technology.
         | 
         | For example the time until the next technology is 1/n, then the
         | technologies per unit time increase exponentially.
         | 
         | This is true within a constant which is
         | https://wikipedia.org/wiki/Euler%27s_constant
         | 
         | A justification for maybe why this is the right model is that
         | the previous n technologies all help make the next technology
         | possible in time ~1/n because they (n existing technologies)
         | all contribute to innovating the next technology.
        
         | dash2 wrote:
         | It would be a change to macro models of economic growth, if
         | accepted. The default is that TFP grows exponentially. There's
         | also a useful comment by Marginal Revolution here (https://marg
         | inalrevolution.com/marginalrevolution/2022/04/ad...) where he
         | points out that TFP itself is really just a residual, and
         | perhaps not a very well-defined concept. That is, if you
         | regress
         | 
         | Y = AK^beta L^(1-beta)
         | 
         | with K being the amount of capital and L the amount of labour,
         | A is just the unexplained "scaling factor" and it gets called
         | TFP. But what TFP actually is... is a bit of a whatever-you-
         | like.
        
           | pishpash wrote:
           | Why is TFP assumed exponential by default? What's the theory
           | behind that? A time lag in the transfer between exponential
           | input growth (e.g. population) and exponential output growth?
        
         | heavenlyblue wrote:
         | How is aircraft limited to the speed of sound still?
        
           | leetcrew wrote:
           | they're not? it just turns out that the most efficient
           | cruising airspeed for passenger jets is in the upper end of
           | the subsonic range. most people can't afford supersonic
           | travel, and the ones that can overwhelming prefer to spend
           | the extra money on a more comfortable subsonic flight.
           | 
           | interestingly, extremely high speeds turn out not to be that
           | useful for military aircraft either. the fastest production
           | fighter jet of all time, the mig-25, was introduced in 1970.
           | the f-22 has a considerably lower top speed, and it's not
           | because they lacked the funding to make it go faster.
        
       ___________________________________________________________________
       (page generated 2022-04-20 23:01 UTC)