Wednesday, March 20, 2019

Thomas Friedman in coversation with James Manyika

The challenges of the modern economy and difficulty of policy building policy responses.

Monday, September 17, 2018

The innovation narrative

Whenever we talk about innovation or innovation policy we are typically speaking within an established narrative. That narrative is we need more innovation. But we never talk about the systems we already have that we are not maintaining or what changes we should prioritise. This talk by Lee Vinsel nicely deconstructs parts of the innovation narrative.









Thursday, March 9, 2017

Economics of Digital Age 5: Technologies of the world unite

While the singularity focusses on AI it is a distraction from what is clearly happening every day at the moment and which wasn't present even a decade ago. That is a technological advance in one place is enabling technological advance somewhere else.

'Innovation systems' has now been replaced by 'innovation ecosystems' in the academic literature. Do we really know more - not really it is mostly just an exercise in talking about ' innovation systems' sounds dated so 70s. I am going to be pointed - I truly do not believe you can have an innovation ecosystem - you can have technology ecosystems but not innovation ecosystems in the geographic sense it is most often used. Clusters is also out of fashion, so we can't use that term either.

The point of ecosystems if anybody bothered to read the ecology literature is that they are so multidimensional. There are population dynamics, energy cycles nutrient cycles food web dynamics, shock and resilience dynamics. When it is used in innovation circles it typically means "do we have the right institutional blocks' in place.

If we were really interested in ecosystems we would be trying to figure out how AI is fuelling autonomous cars and how autonomous cars will rewire cities. Now to be fair this is happening - MIT labs and others have been doing some incredible modelling of possible scenarios with autonomous cars. But that is because it is such an obvious example - the problem is we need more and we need to think how to build an idea of the whole - how do different technologies possibly interact at the economic level.

What is clearly coming is multiple waves of change crashing in on us almost simultaneously. However, publishing cycles remain slow and policy makers are trained in ways that are now decades out of date.

It gets mighty complicated. In a recent article The Economist pointed out that http://www.economist.com/news/briefing/21717365-wind-and-solar-energy-are-disrupting-century-old-model-providing-electricity-what-will  while renewable energy is coming online there are design problesm with the grid itself and as renewables drive down cost it gets less attractive to build the new grids. If you are temped to blow this off as "The Economist" then you should read this by the ABC in Australia on the South Australia power grid situation.

 So while the electricity grids are not built for significant amounts of renewable power, electricity demand will only increase because of new technologies increasing the viability of using batteries/ electricity in an ever wider range of products (IoT, cars, trucks, drones, robots etc etc.)


Monday, January 23, 2017

Economics of Digital Age 4: Measuring the Technology Transformed Economy

Of late I have been struggling to come to terms with the modern technology economy.


A bit of a history lesson



Back when I started in this game there was data on:


  • R&D by business, governments, universities and non-profits
  • R&D personnel by same classification
  • Patent data
  • Publication data


....and then we would mine economic structure data to create information on trade patterns and clusters etc. This is what I have specialised in for 25 plus years. All of the data was supplied by official agencies.

But over the last three years the landscape, it seems to me, has changed dramatically. Back in the 1990s, if we looked at high technology industries (high R&D intensive activities) aerospace rolled out a new aircraft every decade or so and pharmaceuticals companies were and still are notoriously slow to move products from research to market.

IT/computers and mobile phones was thus the sexiest thing to watch - and the number of articles reflects this. Search Research Policy for these terms and any other phrase you care to think of and the results will not surprise you

This is all very fine up until about 2008. The IPhone came out in 2007, Uber launched in 2009. The autonomous mining equipment in Australia, that I have already posted about elsewhere, really started to become a serious game changer in 2008.

Today, there is AI, block chain, big data, translation software etc etc.

I've compiled the following list.



The OECD has published this:




What we are seeing is a revolution in technology that is changing the very structure of the economy. These are not always obvious changes. For example renewable energy is beginning to make a different and contribute to energy grids around the world. Consumers can become producers of energy in smart micro-energy based grids.

It seems that our national accounts based view of economic activity just cannot take account of what is happening.


So What do I suggest?

I think we need to focus data collection and analysis, at least for the near future, on three aspects of the thing formerly called the economy.

1. The tech itself
2. the companies
3. skills, employment and income

The tech

For decades now we have ignored the development of actual technologies, preferring instead to think of 'innovation'. The result is today while there is avalanche of stuff coming, or as Richard Florida tweeted in 2016 a series of 'nested transformations' - for the most part we are all at sea because we don't know what is going on. 

So therefore, let us go back to basics and make things concrete again - let's analysis and compile technology statistics and work out how to compare them. So for example how fast are solar technology efficiency rates changing, how fast is new load capacity being added and what share of the total is this.

Autonomous cares - how fast are they being developed and as soon as they hit the road what is their share vs the installed base share - as we might call it.

etc.

The companies

I think it is clear now that the new mega technology companies are something out of science fiction - we need better data on these companies. But we also need better data on organisational platforms of the economy.  I do not think this needs further justification, it is obvious.

People

The trouble we are seeing in the world with Brexit and Trump reminds us that people matter. We have been so focused on GDP and industries we have forgotten that humanity still craves a "good life". A sense of worth and value, a sense of contributing to the whole.

Take that away and there will again be mass conflict. We must return to being interested in the lives of real people.

Summary

Please stop talking about innovation and return again to focus on what changes are happening to whom and on what scale. Policy frameworks depend upon it.


Wednesday, June 22, 2016

Economics of Digital Age 3: Dis-intermediation & Re-intermediation


The term dis-intermediation gets banded around alot in this early cyber revolution period but what does it mean and how can you show it.


So Wikipedia has the following rather limited definition of intermediation.
Intermediation involves the "matching" of lenders with savings to borrowers who need money by an agent or third party, such as a bank.[1]
If this matching is successful, the lender obtains a positive rate of return, the borrower receives a return for risk taking and entrepreneurship and the banker receives a return for making the successful match.[1] If the borrower's speculative play with the funds provided by the bank does not pay off, the bank can face significant losses on its loan portfolio,[1]and if the bank fails its depositors can lose some of their money if the deposits are not insured by a third party.
In fact with vision, and vision is needed here we need to expand the concept of intermediation to just about any scenario you can think of where there is an actor interposed between step 1 and step 3. So ingrained in us are conventional concepts of human work - most scenarios will not even look like intermediation - they only become so when digital technologies offer new possibilities.

So let us illustrate the concept.

Figure 1.




This gives us a basic template to see what happens when we change the dynamics of the economy.
At the highest level we can observe what happens to entire 'industry complexes' when a digital substitute marketplace is created.

Figure 2.


Pre-2000 there was a substantial industry constructed around music - agents, labels, studios, physical media manufacturing (records, tapes and then CDs) and then the retailers.

However, along came napster and later iTunes and what have we got left.

Fig 3.


What we see here is that Apple iTunes has dis-intermediated the music industry and re-intermediated music as well. It now offers a direct route for artists and traditional music industry operations must go through it to get to the customers. This process has dramatically shrunk the tradition commercial music complex.

Now we can apply this to any example. For example a worker driving a truck. Pre-2000 this idea that the driver is an intermediation point would be treated as complete nonsense - now not so much.

Fig 4.



Fig 5.

We can now move to a scenario when mining companies can dis-intermediate this value chain.


This may look dramatic and that is the point. This logic of dis-intermediation and re-intermediation is intrinsic to digital technologies and we had better start designing data around this concept to understand it better.


Monday, April 11, 2016

Economics of the Digital Age 1 - maintenance

Just some musings today of future costs rather than benefits.

So as a follower of innovation studies and also a collector of serious attempts to look into the future it seems to me that one of the serious flaws in the way we bean count is to only focus on growth. It always seems that studies overshoot the future partly for lack of technical advance but partly because they undervalue the economics of the system. These dimensions are inextrricably link - make enough technical advance and the economics improve - however perhaps the upfront costs of that technical advance are too high to begin with.

It seems to me that in understanding this conundrum our national accounts have categories which are next to useless in the modern economy.

So for example I have often wondered about the accumulated techno-system in societies. So inspired by the great read https://aeon.co/essays/innovation-is-overvalued-maintenance-often-matters-more
here are a few jottings.

That blog immediately triggered so many tangential thoughts. Assimov's comments in the Foundation trilogy that as the Galactic Empire slowly collapsed it could not longer maintain its technologies.

Another was a throw away comment in the current U.S. 'great wall' debate about the cost of maintaining the wall. http://www.cnbc.com/2015/10/09/this-is-what-trumps-border-wall-could-cost-us.html includes an estimate of $750m - per annum for maintenance. If the wall cost $12b even without accounting for inflation you are paying for the equivalent of a new wall every 16 years.

Out of sheer curiosity I looked up the accounts of the local public transport body and it was hard to discern what their actual final maintenance costs are.

But let's just assume that gradually over time not only are our societies accumulating more infrastructure but that as technology is more fully embedded the costs rise.

How to read the image. As we rarely discard technologies entirely or when we do, we replace them with ever more complicated ones. So we add to the techno-system and rarely take away. If you read the chart vertically you can read it that past structures are past onto future generations. Roads build pre-1950 are largely still in use today as entities and locations although they have been upgraded in quality significantly. Therefore the pre-1950s column goes vertically upwards. In the pre-millennial period we started introduced significantly new and varied technologies and we currently stand on the cusp of IOT of things. The graph tries to take account of some efficiencies and productivity improvements from gains from embedded technology - but there will still be more of it. 

So over time the costs of maintenance rise - but it is difficult to determine whether this is proportionate or disproportionate to the rest of the economy. We know that in the USA it is failing. In Europe where taxes are higher infrastructure maintenance is better.

Of course early on councils and governments will embed sensors on everything from water mains to roads - but here is the question - what will break first the sensor or the pipe. If it is the pipe how quickly will cash strapped councils replace the sensor.

With the current obsession with new technology - we only study the system around the production of the new technology - not the technology system itself. Oh and BTW costs are only one element - you actually have to have a techno-system that values education for the skills necessary to maintain the societal infrastructure.