I have an an article in this month’s Communications of the ACM (a leading journal for the computer science community). The main point of the article is that most of the work required by the economy is on some level routine and predicatable. This is true of both blue and white collar work, and it includes many skilled occupations. A key point is that acquiring more skills will not necessarily be a defense against advancing automation technology. Machines are getting better and better at acquiring skills.
You can read the article here:
“Could Artificial Intelligence Create an Unemployment Crisis?”
There is also an introduction by the editor-in-chief, Moshe Vardi, here:
5 thoughts on “Could Robots and Automation cause an Unemployment Crisis — CACM Article”
We will need a new economic system; monthly social credit, interest free loans and banking etc. if regular folk are to make it.
By the time a student who is currently a college freshman today graduates as either a doctor, lawyer, or MBA at the start of the 2020s IBM’s Watson machine will be approximately 15 to 30 times more powerful than the current version is today.
And IBM has already long been making plans that Watson will be taking those six-figure jobs away for a fraction of the cost:
In other words we have already passed the threshold point where it makes no economic sense to pursue a professional education because without that level of income, ($300K a year radiologist jobs will no longer exist the 2020s), it will be impossible for the vast majority of students to ever be able to repay their six-figure student loans.
From here on in the only value that a higher education will have is “entertainment value” and really nothing else.
Nice. All of it: the blog, the article, the tone, …
One might think that those below some level (Harvard’s slice?) would not have any role to play in the coming world. However, all sorts of talents lurk that have not been measured (actually, not even seen). The web (however it’s characterized) allows us a glimmer of a framework for potential contribution by the non-elite on a continual basis.
Too, though, one could even say requirement for such dawns (suggesting, of course, that feature-driven progress (regress, many times) is inadequate). The total costs of systems (the gamut) have not been included within the model, given how it has evolved since the inception of advanced technology. In a sense, there ought to be as much effort spent in checking as there is in computing (even with some consideration of error-correction, and more).
And, machine-learning results (the culmination of such) will deserve scrutiny, to boot.
To be discussed, of course, but, the need suggests a whole lot of employment. Jobs? We’ll need to change that whole concept, so that even those below the 50% level (as mentioned in the article) would participate. After all, that particular measurement is against known tendencies (with biases that need serious examination).
Here’s one oversight of the current approach. That is, in the modern view, debt loads that extend to several levels of future generations is accepted as necessary and natural. A consumer-driven economy would require that the consumer would have the means with which to consume without stacking indebtedness upon her/his progeny.
In short, one might say that the bias of numeracy is one of the worst with which the future economy will have to deal.
This article should be an addendum to his book. As our information technology increases, these jobs will in effect, become worthless. As an American, I am more worried about our economy than the rest of the world. This may sound a little selfish, but it is the world I live in.
From my point of view, it seems the capitalists, in whatever form you want to view them from, (think Corporate Elite) fail to appreciate any type of long term thought on this subject. They have forgotten history, where in the past the Elite were the Monarchies, which have long been destroyed in most parts of the world because of their short term thinking and callous behavior for the society in which they ruled.
These Elite are using their power and influence in the US and Europe to corrupt our societies elected governments into doing their bidding of this continued short term thinking. The Roddenbery Universe, which I see this authors vision to be, came at great cost to humanity before the enlightenment for which we could overcome this trend. If we the People want to ensure that we continue to have a capitalistic economic future, it will require us to act now, before the masses of the 50% in martin’s article decide to turn on the other half of society out of jealousy and vengeance.
As I understand it, the three pillars of capitalism are Land, Labor, and capital. In order for our economy as we know today to continue, these foundations have to be stabilized. Our Elite are doing anything and everything in their power to remove one of these pillars: Labor. Without labor, or more specifically, the Value of Labor, there will be no economy. The only way to put value into labor again is to create scarcity.
Our government and Elite are doing anything they can in the US to increase the supply of labor. And they are doing on both of the fronts that Martin points out; H1 Visas for highly skilled labor, and loose immigration laws, like those in debate right now in Congress. Add to this the fact that they want to provide amnesty to those who are here illegally, and make it more convenient for corporations to take advantage of undocumented workers, the Value of Labor (of which is already declining because of automation, globalization, outsourcing, and innovation) will continue to be devalued to the detriment of our economy and our society.
What truly needs to be done in the US, which at the moment is the largest in the world, is the People need to take back control of their Government, and reduce the influence the Elite have. This will be a monumental effort, but it starts there.
Great article. I don’t think that robots will cause any problems until they can be made at a pace faster than the economy can accommodate them. Just think about how long it takes to make a robot. It has to designed, debugged, prototyped, built, tested. It takes for-freakin’-ever! When news about robots come out the stock market reacts cause it predicts events months in advance, so it’s like we are prepared for it. But when robots start making robots…. they’re gonna get made much faster. That’s all I’m saying, meatball.