3 ms·

The author doesn't seem to include any metrics for how they're measuring technological changes, or am I missing that? Seems like a graph and numbers would be us

by mlb_hn 7y ago

The author doesn't seem to include any metrics for how they're measuring technological changes, or am I missing that? Seems like a graph and numbers would be useful for measuring what sort of slowdown it's talking about; most of the arguments seems to be looking at what wasn't accomplished.

davidivadavid 7y ago
That's the main problem with those articles. We have no metric for innovation. So all we get is a bunch of cherry picking of what constitutes "major" "progress" based on undefined criteria, and then comparing those things across different eras. Some organizations have created innovation indices, but they tend to measure inputs (e.g. R&D spending) more than outputs, or measure outputs that have questionable relationships to innovation (e.g. number of patents registered).
- beefman 7y ago
  There are some obvious metrics that are likely to be very good: global energy use, global GDP, and economic efficiency from their quotient. It's the data we're missing (prior to 1970). It could be estimated, but almost no one has tried. Another idea is to count the occurrences of strings representing years ("1974" etc) on wikipedia.... possibly weighted by the PageRanks of the containing articles...
  - davidivadavid 7y ago
    I don't know about "very good." GDP is sometimes underrated as a metric, but it does have serious flaws that would prevent me from calling it "very good." Capabilities-based approaches seemed more promising for a while, but it seems like they haven't resulted in anything better yet. I'm still unclear why there seems to be so little research on measuring actual value created by measuring things like consumer surplus (at least heuristically).
    - beefman 7y ago
      Really? I'd call it 'more descriptive of human civilization than any scalar has a right to be'. What do you make of the S&P 500 index as a descriptor of the stock market? Consumer surplus is one of those economics concepts that probably doesn't exist. People will generally pay anything they can afford. Fortunately, what they pay for positional goods and such mostly washes out as transfers or asset inflation and doesn't pollute GDP, which measures intrinsic value (bits, ultimately).
      - davidivadavid 7y ago
        Sure, it could be one of those things that's "the worst except for all the other options." It's still flawed, the flaws have been known for a long time, and their magnitude is potentially large enough to warrant looking for other indicators. The S&P 500 describes the S&P 500 fine, but it seems like a much simpler problem? They represent an estimate of a future cash flow generated by a basket of companies. That's way more straightforward than measuring value perceived by consumers. How does consumer surplus not exist? How would you value something like e.g. Wikipedia?
        beefman 7y ago
        Stock market and GDP indices are generally the same in that they are sums of correlated time series. Summing suppresses uncorrelated variance (often measurement noise) and retains correlated variance. The index's returns are essentially the first principal component of the returns of the index memebrs. This is how the CAPM works – stocks are analyzed by factoring out their responsiveness to the market. Stock market and GDP indices are also particularly related, as stock prices are tied to corporate earnings and corporate earnings are a major component of GDP. GDP has nothing to do with human perception. It measures the productive capacity of civilization, which is a physical quantity.[1] It's often claimed that if you give something like Wikipedia away for free, its value will not show up in GDP. In fact, if Wikipedia is valuable, it must have a positive impact on civilization. This will be reflected in GDP, even if it's in the form of extra time/money people spend on ice cream in the summertime. The maximum someone would pay to access to Wikipedia is likewise irrelevant. For one thing, it's a function of their personal wealth, which is an unrelated quantity. But in any case, paying more than something costs to produce just increases the buying power of the people selling it... But it can't increase their buying power above what civilization can produce. It's a just a wealth transfer. Money can't defeat physics. All of that is straightforward. What can be surprising (at first) is how well GDP measures human happiness,[2] drives wealth and income inequality,[2] and predicts election outcomes.[3][4][5] Economic peaks coincide with moon landings, complexity in popular music, the discovery of canonical physics, commercialization of computing technology, and various athletic records.[6] [1] http://www.inscc.utah.edu/~tgarrett/Economics/ http://www.inscc.utah.edu/~tgarrett/Economics/ [2] http://lumma.org/econ/Happiness.html http://lumma.org/econ/Happiness.html [3] https://doi.org/10.1016/j.electstud.2014.05.002 https://doi.org/10.1016/j.electstud.2014.05.002 [4] https://pollyvote.com/en/components/econometric-models/time-for-change-model/ https://pollyvote.com/en/components/econometric-models/time-... [5] https://fivethirtyeight.com/features/which-economic-indicators-best-predict-presidential-elections/ https://fivethirtyeight.com/features/which-economic-indicato... [6] https://gist.github.com/clumma/90a1a53a1376699dd4001823e2760904 https://gist.github.com/clumma/90a1a53a1376699dd4001823e2760...

by mlb_hn 7y ago

davidivadavid 7y ago

That's the main problem with those articles. We have no metric for innovation. So all we get is a bunch of cherry picking of what constitutes "major" "progress" based on undefined criteria, and then comparing those things across different eras. Some organizations have created innovation indices, but they tend to measure inputs (e.g. R&D spending) more than outputs, or measure outputs that have questionable relationships to innovation (e.g. number of patents registered).

beefman 7y ago
There are some obvious metrics that are likely to be very good: global energy use, global GDP, and economic efficiency from their quotient. It's the data we're missing (prior to 1970). It could be estimated, but almost no one has tried. Another idea is to count the occurrences of strings representing years ("1974" etc) on wikipedia.... possibly weighted by the PageRanks of the containing articles...
- davidivadavid 7y ago
  I don't know about "very good." GDP is sometimes underrated as a metric, but it does have serious flaws that would prevent me from calling it "very good." Capabilities-based approaches seemed more promising for a while, but it seems like they haven't resulted in anything better yet. I'm still unclear why there seems to be so little research on measuring actual value created by measuring things like consumer surplus (at least heuristically).
  - beefman 7y ago
    Really? I'd call it 'more descriptive of human civilization than any scalar has a right to be'. What do you make of the S&P 500 index as a descriptor of the stock market? Consumer surplus is one of those economics concepts that probably doesn't exist. People will generally pay anything they can afford. Fortunately, what they pay for positional goods and such mostly washes out as transfers or asset inflation and doesn't pollute GDP, which measures intrinsic value (bits, ultimately).
    - davidivadavid 7y ago
      Sure, it could be one of those things that's "the worst except for all the other options." It's still flawed, the flaws have been known for a long time, and their magnitude is potentially large enough to warrant looking for other indicators. The S&P 500 describes the S&P 500 fine, but it seems like a much simpler problem? They represent an estimate of a future cash flow generated by a basket of companies. That's way more straightforward than measuring value perceived by consumers. How does consumer surplus not exist? How would you value something like e.g. Wikipedia?
      - beefman 7y ago
        Stock market and GDP indices are generally the same in that they are sums of correlated time series. Summing suppresses uncorrelated variance (often measurement noise) and retains correlated variance. The index's returns are essentially the first principal component of the returns of the index memebrs. This is how the CAPM works – stocks are analyzed by factoring out their responsiveness to the market. Stock market and GDP indices are also particularly related, as stock prices are tied to corporate earnings and corporate earnings are a major component of GDP. GDP has nothing to do with human perception. It measures the productive capacity of civilization, which is a physical quantity.[1] It's often claimed that if you give something like Wikipedia away for free, its value will not show up in GDP. In fact, if Wikipedia is valuable, it must have a positive impact on civilization. This will be reflected in GDP, even if it's in the form of extra time/money people spend on ice cream in the summertime. The maximum someone would pay to access to Wikipedia is likewise irrelevant. For one thing, it's a function of their personal wealth, which is an unrelated quantity. But in any case, paying more than something costs to produce just increases the buying power of the people selling it... But it can't increase their buying power above what civilization can produce. It's a just a wealth transfer. Money can't defeat physics. All of that is straightforward. What can be surprising (at first) is how well GDP measures human happiness,[2] drives wealth and income inequality,[2] and predicts election outcomes.[3][4][5] Economic peaks coincide with moon landings, complexity in popular music, the discovery of canonical physics, commercialization of computing technology, and various athletic records.[6] [1] http://www.inscc.utah.edu/~tgarrett/Economics/ http://www.inscc.utah.edu/~tgarrett/Economics/ [2] http://lumma.org/econ/Happiness.html http://lumma.org/econ/Happiness.html [3] https://doi.org/10.1016/j.electstud.2014.05.002 https://doi.org/10.1016/j.electstud.2014.05.002 [4] https://pollyvote.com/en/components/econometric-models/time-for-change-model/ https://pollyvote.com/en/components/econometric-models/time-... [5] https://fivethirtyeight.com/features/which-economic-indicators-best-predict-presidential-elections/ https://fivethirtyeight.com/features/which-economic-indicato... [6] https://gist.github.com/clumma/90a1a53a1376699dd4001823e2760904 https://gist.github.com/clumma/90a1a53a1376699dd4001823e2760...