⬆️ ⬇️

Italian robot strike

The story of how the development of automation will lead to the collapse of a civilization known to us. Maybe.



In our blog, we already wrote about some of the consequences of automation in the article " Automate it ." Here I will try to talk about the ideas of Martin Ford and his book “ Rise of the Robots ”. Ford wonders what we, as humanity as a whole, will lead to the further development of robotization or, more generally, automation, where most of the work will be done by mechanisms, robots and computers.





')

Isaac Asimov is not least known for the Three Laws of Robotics . The robot can not harm a person by action or inaction, must obey the orders of the person and ensure their own safety.



In turn, the Italian strike is called a form of protest, in which employees are extremely strictly adhere to official duties and instructions, that is, behave like robots.



The modern world is full of examples of automation. Perhaps now you are reading the article from your mobile phone, sitting on the couch, only a few hours after it was published. But thirty years ago, to achieve the same result, the article needed to be folded, printed, distributed into kiosks, and you would have to buy a magazine or newspaper, bring it home and only then read while sitting on the couch. But we are running ahead.



Muscle automation



One of the relatively recent examples of automation is the use of steam engines. Machine labor replaced horse labor, hence the unit of measurement - horsepower.



Mechanization has fundamentally changed the approach to farming: the tractors freed up a lot of workers who had previously been engaged in agricultural production, which led to urbanization.



Mass production, in turn, was not limited to the introduction of a conveyor approach at the beginning of the 20th century. By the end of the last century, industrial robots were performing more and more operations on the conveyor. For example, in many automotive industries, bodywork, from stamping to painting, is a fully automatic process. However, people are still involved in the last stages of painting.



Automation and Machine Learning



But these are trivial examples, Ford cites more relevant and curious types of automation. Robots that produce sushi or burgers that go straight from the conveyor belt to your plate, and no dirty hands along the way. Self Driving Trucks threaten to stop the work of truckers, Uber is asleep and sees how to introduce taxis everywhere that do not require drivers, and Amazon plans to release an army of delivery drones, and so on.



The essence of modern automation is not to optimize a person’s work, but to completely eliminate him from the production cycle, leave him out of work, replace him with a robot.



Back in 1947, Norbert Wiener saw the far prospect of using computers to solve control problems - cybernetics. So distant that in the next 50 years no significant results were obtained in this direction. Cybernetics fell into disgrace, and as a result, we now use the term Machine Learning, although the essence remains the same.



Modern advances in machine learning algorithms, such as the victory of Alpha Go, show a qualitative leap in the development of this application field. Previously, automation and robotization competed with a person only on the labor market of physical strength: those who quickly dig a hole, transport more tons of cargo, drill more holes. We were fully confident that intelligence is what distinguishes us from robots. Let them be faster and stronger, but where it is necessary to apply sharpness or make a decision, they cannot take their work away from us. On this topic, you can watch the video from TED: " The works that we give way to machines ", 2016. The bottom line is that any intellectual work that consists in making a decision based on a typical set of factors can be automated by machine learning algorithms if we have enough examples to learn from them. Moreover, the result will be even better.



The fact that automation and robotization deprives us of jobs is understood by Luddity in 1812, but by and large, society as a whole always had a place to use the freed up human resources. In the end, there is employment in the service sector. But what will happen if large online stores bring the rails to each house, along which the ordered goods will be delivered within ten minutes?



Total Capitalism



The less obvious and much more interesting aspect of total automation, shown by Martin Ford, is to concentrate profits and influence among a small number of people.



Automation inevitably leads to increased productivity. If earlier for work in port hundreds of loaders were necessary, then for service of automatic cranes only ten people can be required. And the arrived cargo is delivered by container ships, in the crew of which not even five will be gathered. At some IKEA plants, furniture production is already automated to such an extent that between the driver who delivered the raw materials to the plant and the driver who picks up the finished product, no one directly participates in the production process. Perhaps it is no coincidence that these are the production IKEA left in Sweden. And we remember that the work of truckers is already under threat from the robots.



The picture of the future, drawn by Ford, is not only gloomy, but realistic.



Complete automation will not only lead to the loss of work for the majority of the population, but also to a concentration of profits among the overwhelming minority. And last but not least we, the developers, are doing everything possible to achieve this result. Perhaps the basic income will help offset unemployment, but this idea also has obvious shortcomings.



Automation Risks



Moreover, total automation, excluding a person from the production process, inevitably leads to stagnation. Once a configured process continues to work, but ceases to improve. The Driving Honda book provides an example of how full automation of production and assembly of a part leads to regression, when improvements in the production cycle cease to be made due to the fact that the operation is automated and its change is too expensive, which does not happen with manual operations.



For example, do not need to go far. I think many of the developers have come across a situation where a module in the system works quite well and there is no need to make changes to it. Perfect automation. But when three years later someone still needs to make a change in it, the question naturally arises - how does it work?



One of the options for the death of mankind Azimov believed the loss of knowledge about technology. But perhaps it will not take millions of years to forget everything: the key technology forgotten can be not the design of a nuclear reactor, but the principles of growing potatoes.

Source: https://habr.com/ru/post/344608/



All Articles