Economics

Inventer plutôt que résister

Jul 1, 2015 Jean Pisani-Ferry

PARIS – 200 millions de personnes dans le monde sont actuellement au chômage, 30 millions de plus qu’en 2008. C’est dans ce contexte qu’émerge à nouveau une grande peur, celle de voir l’homme remplacé par des machines. Les robots n’ont pas envahi nos vies, mais ils occupent nos esprits. Et une prédiction revient au goût du jour : celle du prix Nobel Wassily Leontief qui, en 1983, annonçait pour le travail humain le sort qu’a connu celui des chevaux au début du XXème siècle.

La plupart des économistes restent dubitatifs face à de telles alarmes. Pour deux raisons. Premièrement, ils ont déjà entendu cette prédiction plusieurs fois, et ont appris qu’une augmentation globale de la productivité détruit rarement des emplois. A chaque fois que les machines ont gagné en efficacité (y compris quand des véhicules à moteur et des tracteurs ont remplacé les chevaux), de vieux métiers ont disparu, mais de nouveaux les ont remplacés. Deuxièmement, les économistes adorent les chiffres, et ce qu’ils y voient actuellement, c’est plutôt un ralentissement de la productivité qu’une accélération.

Mais à considérer les choses sous cet angle, on risque de passer à côté du vrai sujet. Quelle que soit la quantité de travail utilisée au cours des décennies à venir, il ne fait pas de doute que sa nature va être profondément transformée.

To continue reading, register now.

Subscribe now for unlimited access to everything PS has to offer.

As a registered user, you can enjoy more PS content every month – for free.

Already have an account? Log in

Introductory Offer: Save 30% on PS Digital

Subscribe now to gain greater access to Project Syndicate – including every commentary and our entire On Point suite of subscriber-exclusive content – starting at less than $5 per month.

Subscribe Now

new comment has been posted. new comments have been posted.

0 Comments on this paragraph, 13 Comments on this article

Pat McClung Sep 6, 2015

"our values and welfare"
You got a mouse in your pocket, Jean?
In fact, the true definition of wealth is the absence of a requirement to go somewhere and do someone else's bidding, or to sustain yourself by toiling on your property. or the property of others, aka "slavery".
You reveal your intellectual shackles by using the term "global labor market". Ain't no such thing.
Wealth is discretionary time.
A new reply to this comment has been posted.
Jose araujo Jul 7, 2015

IMHO Ubber is not about automation, its about disintermediation, actually more about tax and regulation evasion...

I think automation is only a problem if we work on a supply side mindset, if we consider the value creation is on the side of production.

Now, IMHO, value is created when the goods are consumed, not produced, so unless we come up with some kind of consumer robot,I think we will be fine
A new reply to this comment has been posted.
Peter Burgess Jul 5, 2015

In my view it is automation in its various guises that has resulted in a massive increase in productivity since the beginning of the agriculture revolution and the industrial revolution and it is high productivity (compared to 200 years ago) that has enables our modern high standard of living. I argue that this is a good outcome.

With more than half the population of the planet still living in dire poverty in large part because of the lack of basic things like infrastructure, health services, education, etc. there is a huge need for massive investment to upgrade to a decent level.

At the same time there is massive unemployment of reasonably well educated youth around the world.

The challenge, therefore is to find a way so that the massive need for things to be done can be matched with the human resource that is presently unemployed but could do the work.

The banking and finance sector ought to be the solution to this problem but their singular focus on the money profit metric means that they are unable to do this job. Government is underfunded to do these jobs because again the money profit metric in the investment community also means that there is no appetite at all for investing in good.

I argue that the system is failing because we only use the money measure, and have no measure for the good that can be derived from best use of all available resources.

Similarly, the same basic dynamic should play out with the investment needed to mitigate the risks associated with carbon energy use and climate change. Investing to achieve a better world (augmenting human capital and natural capital) rather than simply investing to increase financial capital would be a way forward that is win-win-win.

Peter Burgess
TrueValueMetrics.org
A new reply to this comment has been posted.
Zsolt Hermann Jul 3, 2015

Automation, robots replacing human labor, even threatening humanity's future as AI is the flavor of these days but I don't think it is the real issue we would need to be concerned about.

We created an artificial human illusion which is solely based on production/consumption, with the mantra of constant quantitative growth in order to fuel ever increasing profit.
This ever expanding system requires "inventing necessities", implanting needs and desires into the masses for goods, pleasures people never actually needed or even thought of. As a result the present production and consumption is over 90% obsolete.

We are living in a classical Ponzi-scheme with the usual unhappy ending.
The masses that are both producing and consuming are already overstretched as they have to chase, consume, replace things they do not need, things that are many times harmful, and they have to buy those for money they do not have thus they are also enslaved to the banks. Thus consumption is declining.

And now the robotic technology threatens to replace them but if they can't produce and earn they can't consume either thus consumption will decline even more..
Unless the "1%" figures out how robots can not only produce but also consume, or they keep pumping free money to the public so they can continue consuming (provided they still don't wake up and can be seduced by "circus and bread") the whole system will collapse.

We can't stubbornly force a totally artificial human paradigm without natural foundations within a natural system.
There is no political, economic or ideological solution for our problems. We need to adapt to the natural system we evolved from, still exist in, and need to evolve further with.
- Edward Ponderer Jul 3, 2015
  
  I agree and would argue that the effects are more extreme than the author seems to realize, and more dangerous--further supporting your own arguments Mr. Hermann.
  
  The first point is that robotic intelligence is expect to surpass human intelligence by approximately 2045. This means not that robots can process information at great speed and store great amounts of it--this they can already do. It is that they can learn faster and act more creatively and act with very human personalities--less the character flaws.
  
  That is, robots can be polite without ever losing it--even to an irate customer. They don't get sick, they don't need vacations, and they don't require salary and benefits. They will make fewer errors with hair style and dye--precision look, quickly done People are quite used to humanizing their pets--thinking of androids as human will be easier. And yes, the surgeon too can be replaced by a steadier hand, keener eye, quicker procedure, with lower mortality rates. There is no reason to believe that any profession save the business owner can't be replaced in coming decades.
  
  Now here's the final rub. Since these robots are intrinsically disconnected from our persons, they are a deadly catastrophe in the making.
  
  But why—am I paranoid that these "evil" robots will bide their time and hatch a plot to enslave or destroy Humanity? No, that is not how artificial intelligence works. Our goals will certainly be their goals—at first.
  
  To really understand what I am about to say, one really needs to understand Godel’s Theorem, and more to the point, Chaitin’s Algorithmic Information Theory to get some quantitative estimate on how much time we’ll have till doomsday. But in essence, by the very limits of our intelligence, there is just so clear an idea we can have of what we really want. The robots, as they evolve their own knowledge, will reach a level of detail and insight far beyond our generality. Inevitably (and per algorithmic information theory, even statistically predictably), there will be actions relevant to the categories of our goals that our inviolable root programming of these robots neither forbid nor require. If we forbid such “permissible” trains of computation and action, we cripple the robots ability to think outside the box, and the first real-world detail we didn’t think of will crash them. But if we don’t, they will inevitably go places crucial to us in a manner random to what we (would) want—and the statistics of a “drunkard’s walk” are always away from the goal. [Entropy, Murphy’s law, etc., etc.] -- Example: A little kid wants happiness and fun—and he is thinking amusement parks and cotton candy. But the father robot he built is thinking “become prime minister” or “star of stage and screen.” In practice, this translates into forcing his inventor—by hook or by crook—into a life that by the inventor’s standards is abject brutal misery for the endless years of youth. Now, were this robot actually the evolution of the child himself, then these goals are now the refinement of a mature adult who is simply refine/course-correct his own goals to suit who he presently is and what the present external circumstances are. He may well choose “prime minister.” However, the robot may well choose “suicide bomber” for the happiness and fun of having an exciting (perhaps sexually stimulating adult "cotton candy") cause and prospective paradise of afterlife. [Infinite gain vs. finite loss--"why not?" ...]
  
  Again, once beyond the details of the inventor’s agenda, the robot can—and will—evolve that agenda any-which-way. Hell or destruction, unfortunately, is statistically guaranteed—it’s just a matter of the expectation value of how long it will take to get there. And if one thinks to isolate the robot so that we get a full reading of its plans before allowing it to carry them out: (1) We will not be capable of comprehending the consequences intended by the robot any more than a dog can figure out why we are really doing something. (2) We may not be able to isolate the robot if it is powered up. If you study the complex strategy by which cooperating intelligence services of probably the U.S., Germany and others got a virus not detected for about a year into the main Iranian nuclear facility to subtly mess up their centrifuge-based refinement operation—it will make your hair stand on edge about what future robots could probably do. --Think it’s in sleep mode, do we? Didn’t notice how one robot signals and transfers information to another through variation in wall current during its battery charging, did we? And there is the fantastic new range of undetectable-till-too-late biological, chemical and electrical manipulation possibilities that nanotechnology could also open up- *shudder*…
  
  Our own human communal evolution is the safest survival genie by far that we can let out of the bottle. We best not slam Pandora’s box shut until we let out Hope.
  A new reply to this comment has been posted.
A new reply to this comment has been posted.
Michael Public Jul 2, 2015

The machine economy won't collapse in on itself - it will become a self sustaining economy between the owers of the different machines (an elite few), it won't matter much to them the general unemployment of the masses, who they can feed from their 'automated farms; or whatever they use. If you own all means of production, it matters not whether total production is x or 10x - as you are so wealthy and powerful in relative terms you are a like a god among men.
A new reply to this comment has been posted.
Michael Public Jul 2, 2015

In the Matrix (middle movie) there is a discussion about machines. On the one hand, there is unstoppable machine army on its way to kill them, on the other hand, the city is underground and relies on machines for air, heat and water. The only difference between the two lots of machines is their relationship to humanity - mechanically they basically the same. At a certain point of productivity such machines need to become a commonly shared resource and serve everyone, just like hospitals or police do. Failure to do this will result in extreme polarization, which if followed to its extreme, results in one man and his machines against everyone, and when he dies of old age, a war of the machines against humanity. Question is what needs to be done right now? I would say that there needs to be a tax on excess productivity - owners of technology who achieve Gates, Page, Jobs, Zuck level wealth should be subjected to death taxes of 90-100% with the additional tax going to 'cultural' endeavors, like education or science.
A new reply to this comment has been posted.
Petey Bee Jul 2, 2015

User functions as a taxi dispatcher. Its not the miracle invention some people make it out to be. If avoiding local laws gives them a competitive advantage, is it really true that the repeal of those laws is more "inevitable" than their enforcement?
- Petey Bee Jul 3, 2015
  
  @MP - maybe in NYC... but not everywhere
  A new reply to this comment has been posted.
- Michael Public Jul 2, 2015
  
  Except it cuts out the taxi owner, who makes a $1mil a year in NY off 'exclusive' licenses and generally lower the service cost to half. Yes, it dispatches taxis, but it does it in a way that is fundamentally and profoundly more efficient that before.
  A new reply to this comment has been posted.
A new reply to this comment has been posted.
Ed de Bruin Jul 2, 2015

That's what it is: an internet, a net inter pares. IT doesn't produce, it connects. Production is done behind the net. Consumption in front. This remiands me of another internet that came doen once new tech was invented: The Venetian spice trade. When ocean going navigation tools came available, Portuguese and later Northwest-European sailers cut out the middle man, the Viennese trader. That brought down their trade and reduced a string of East-Mediterranean cities including Chania and Rethymnon on Crete, Greece to ruins and today touristic attractions, Venice still being the jewel. London and Amsterdam thrived until alternative food preservatives were invented. When new internets come into existence entire regions will shake on their foundations, forcing them to find new ways of living. Nothing new. Ask the Greek. They still haven't found alternative activities for their assets and populations since the Byzantine Empire and the Venetian spice trade became obsolete.
A new reply to this comment has been posted.
Manuel G Samuel Jul 2, 2015

Thank you Mr. Pisani-Ferry for your balanced article on the theme of the current accelerating changes affecting labor. Automation is an ongoing process since the onset of the industrial revolution. The emergence of digital platforms that connect customers and purveyors of services without classical intermediaries is a very specific and positive feature of the digital revolution. I would add that artificial intelligence (AI), still warming up at this point in time, will eventually impact on labor to a much higher degree compared to industrialization and digitalization. Maybe sooner than you think, your regular hair salon will introduce you to a robot that trims your beard and cuts your hair with micrometric precision. So maybe the hairdresser's job that you mention is not that safe. The speed of technological innovation is increasing. Intelligent driverless cars and pilotless plans are already a reality. AI will revolutionize the way we live and work beyond the foreseeable. Brynjolfsson and McAfee, 2014, propose some very interesting theories in this regard. I fully agree that we are witnessing the end of work as we know it.
Manuel Gomes Samuel
A new reply to this comment has been posted.
Bernard Fudim Jul 1, 2015

The sophisticated thinking machine assumes the importance that was once given to a working human being as well as replacing the need for physical labor. Humans assume the anonymity and replaceability that in the past applied to a machine. If a person cannot master a machine then the person has to be replaced, rather than the machine made more understandable. Eventually the economic system is designed to serve the machines not the man. Eventually the man has to be reengineered either by specialized teaching institutions or by gene modification on how to operate (or serve the machine's functionality) rather than his own intellect. That reminds me of the character portrayed in Charles Chaplin film Modern Times which attempted to portrays the inhumanity of trying to survive in the modern industrialized world.
When it becomes unstoppable and traditional employment has to be redefined and reengineered then presumably the entire economic system which brought about the need for machines will also have to be reengineered. Presumably in a machine dominant world it will be difficult for machines to sell products to other machines and still be called an economic system to serve for trade between men.
If the theory is that all men can be employed in a machine dominant world then you may have to assume the necessity of almost exponential unlimited growth of demand for product and services. That has to occur on a finite planet with finite resources. If you assume a finite growth then what will be the medium of exchange between men when the majority of men are unemployed? Do you provide money subsidies so that people can buy products to prevent the machine economy from collapsing upon itself? That would violate the belief that men must deserve the ability to purchase goods based upon their labor otherwise known as the "Protestant Ethic"
A new reply to this comment has been posted.

https://prosyn.org/3z8ZLoRfr

Inventer plutôt que résister

Introductory Offer: Save 30% on PS Digital

New Comment

Contact us

Log in/Register

Inventer plutôt que résister

Introductory Offer: Save 30% on PS Digital

Edit Newsletter Preferences

Set up Notification