Tuesday, October 27, 2009

What happens if we automate everything

According to legend, if you can boil a frog if you put it in cold water and raise the temperature to boiling slowly enough.

Strangely enough, there is even empirical evidence for the claim. One physiologist, a Freidrich Goltz noted that if you remove the brain from a frog, it will no longer hop out of the water. Small detail that, the brain had been removed. In another source which I was unable to track down immediately, a researcher noted that frogs with intact brains will tolerate higher temperatures if the water temperature is raised slowly. In several hundred observations, every frog jumped out except for the one which became the 'proof' of the anecdote.

The reason I raise the question is that I have been thinking about the question of what happens if computers and robots start working so well that they effectively replace most forms of human labor? The idea is not so far fetched as it was even a decade ago. Twenty years ago practically all leisure time was spent either sleeping or watching television. Today people while away the hours on the Internet.

As one of the instigators of this brave new world we have inflicted on you all, one of the most frequent objections made against proposals I make is that they will 'kill jobs'. To which I answer that eliminating unnecessary jobs is a good thing. One of the biggest challenges we face in the industrialized world is the fact of an aging population. People are living longer and spending a lot more time in retirement. So the people in work are having to work for the people who are retired. To date we have answered this issue by importing labor from the rest of the world. But those countries are also starting to face the same trends and those sources will dry up, if not reverse the emigration trend as ex-patriots start returning.

But what if every job turns out to be replaceable?

Economists are not much help on questions of this type. The best they can provide is an explanation of the circumstances that caused the last event. At worst they attempt to bully 'unqualified' commentators into silence with specious claims of expertise.

Only about 15% of the wholesale price of a book is paid as royalties to the author. If we add together all the costs associated with producing the content of a book they represent less than 10% of the retail price paid by the end customer. The other 90% represents the cost of paper, printing, distribution and retailing. Costs and jobs that will be eliminated entirely as the industry moves from paper to digital.

The story is not new, the introduction of industrial robots has led to a similar transformation in manufacturing industry. Employment in agriculture, already negligible, continues a steady decline. What is different this time is that the job losses are affecting the part of the economy that grew under the previous transformation. The 'knowledge workers' whose skills were meant to guarantee employment turn out to be as replaceable as the Victorian farm workers were.

Perhaps the biggest shock of the current recession was the realization that the feckless idiots funding extravagant lifestyles by borrowing against their house were previously the foundation of our apparent prosperity. Sales dry up when there is nobody with money to buy.

In retrospect, it appears that the introduction of containerized shipping transport is a much more likely cause of the recession in the mid 1970s than the oil price shocks on which it is usually blamed. The patient was already sick, the oil price shock was simply the trigger that sent him to the hospital. If the Wall Street crash had been the sole cause of the great depression it should have ended with the recovery that followed FDR's initial recovery act. Instead the country faced a second dip that I suspect was more likely caused by the displacement of agricultural workers displaced by rural electrification.

It is perhaps the worst legacy of Keynes that economists insist on ignoring the impact of technology on the economy in favor of mathematical and fiscal explanations for economic trends. It is not surprising why they prefer this, you can't plot a smooth curve through the invention of the shipping container, the barcode or the World Wide Web. More importantly, predicting economic trends from technology trends would require intellectual skills that modern economists have largely abandoned in favor of abstract algebra.

So we are caught in something of a trap. The aging population will cause our living standards to collapse if we do not eliminate jobs fast enough. But if we eliminate too many jobs we risk a depression or a slump. There may not even be a 'Goldilocks solution', merely eliminating the unnecessary jobs means that the economy has the additional capacity to meet the increased needs, it does not guarantee that the resources can be moved from one place to another.

We are caught in a trap, but here is where the frog comes in. If the water gets too hot, we can jump out.

As technologists, our only responsibility is to make sure that the economy has sufficient capacity to meet necessary needs. We cannot and should not worry about doing that job too well. If there is a temporary shortage of demand then it is the responsibility of government to address the issue. If we reach the automation limit then society is going to have to invent a new basis for allocating resources.


rajnish said...

. love to see this discussion! It’s great to see you all working through the issues and also, it’s great to see recommendations for testing. In the end, it’s what your actual users do and prefer that should be your biggest driver in making these decision.
web designing

NG said...

The article is a very good; we are a SEO company searchrounds based in India. We have good readers in the team. Who knows the quality of work. We will pass it post to the readers and I know that they enjoyed it too!
seo India

aa said...