With the advent of the industrial revolution, many products were manufactured in large factories employing large numbers of workers. These factories were more productive and profitable due to the large quantities produced in short periods of time because of mechanization. This was in stark contrast to the slow, skillful production of smaller quantities of quality products created by artisans earlier. However, mechanization still involved humans whose productivity increased by working with machines and tools that were externally powered. Hence, mechanization led to the enhancement but not the replacement of human labor. However, since humans do commit errors, have variable output due to requiring breaks, sometimes don’t feel well, have distractions, are emotional, and get bored and dissatisfied while performing repetitive tasks, business owners started looking out for something better than mechanization to further boost productivity and profits. With automation, human labor itself is replaced by machines. Mechanization is represented by a human continuously tightening bolts with a manual wrench or a power tool; automation, on the hand, signifies a machine automatically tightening bolts without human intervention.
Automation changes human work by:
- Enabling customer do-it-yourself
- Substitution of humans with machines
- Substitution while complementing higher skills
- Shifting from physical to cognitive
- Creation of new tasks.
Enabling customer do-it-yourself
More and more technologies are simplifying tasks and enabling people to do things for themselves. For example, digital photographic technology made do-it-yourself possible for customers with photos from digital cameras developed and printed on a home printer, as compared to sending a camera roll from a mechanized camera to a film processing outlet earlier. As a result, the market for printing and developing film has disappeared, and so have the jobs. Also, with the introduction of the customer self-service ATM, jobs were lost to automation with bank tellers and their managers made redundant. Further, tax preparation software, which substituted for trained tax accountants, enabled consumers to quickly and economically file tax returns by themselves from their personal computers or even mobile apps. This made trained tax accountants redundant. More recently, self-checkout kiosks, offered by many large retailers, have displaced proficient checkout cashiers, providing sufficient cost savings for the use of technology. Although, presently a few low-cost human attendants are made available solely for providing guidance to amateur customers in the use of the high-technology self-checkout kiosks, with the self-checkout shopping cart on the horizon, even these support staff may become redundant.
Substitution of humans with machines
When a new technology automates a set of tasks, involving physically demanding, repetitive, and rote activities that were previously done by workers, it substitutes machinery for people. Since the machinery performs these tasks more economically, faster, or better than the workers who previously performed these tasks, this process effectively results in higher aggregate productivity. In the past, for example, sound recording technology in motion pictures substituted for musicians in movie theaters, automation with electric relays displaced elevator operators, and software technologies in airplanes replaced flight engineers. Using human workers in manual toll collection process is slow, causing traffic jams. It has therefore been replaced with automatic tolling, where various detection and payment technologies are used to make collection effective and accurate. This speeds up traffic flow through toll stations, and also reduces pollution. As a result, automatic tolling has seen more and more adoption on highways throughout the world, resulting in redundancy of human toll collectors. Although these processes do raise productivity, the gains through substitution of machines for workers generally are available to firms via higher profits and to customers via lower prices. However, displaced workers and their families usually bear the costs, including their country’s economy, through the social benefit programs implemented for workers that lose jobs.
Substitution while complementing higher skills
Many workplace technologies substitute for one set of tasks while complementing another set of tasks. With these technologies, automation helps raise productivity of some higher skilled jobs, while substituting for the lower skilled ones. For example, mechanized farming displaces agricultural workers but empowers farmers with improved productivity. CAD software substitutes for draftspersons but complements architects and engineers for rapid exploration of design options. By avoiding painstaking drafting of numerous paper drawings, this process enables architects to design more complex buildings faster. Also, medical imaging tools substitute for technicians but boost the speed and accuracy with which medical experts diagnose patients. In another example, computerized telephone agents (bots) deployed by airlines and hotels displace humans servicing routine queries. More complex queries, which these bots are as yet unable to handle, are passed on to humans who are better capable of handling these. With the resulting productivity gains often leading to lower prices, improved quality, or greater convenience, employment of experts performing these tasks may rise with the widening use of these technologies. But, on the other hand, complementary technologies may displace lesser educated workers from clerical, sales, production, and operations occupations towards jobs that require generic skills and offer very low wages.
Shifting from physical to cognitive
Complementary technologies tend to complement the cognitive and creative capabilities of highly-educated professionals and augment productivity in their current job tasks, tending to increase their earnings. They also frequently shift the nature of the work from physical to cognitive, enabling new capabilities in the process. Automation raises the value of human expertise in developing and guiding complex production processes. It also provides workers with tools to easily convert their ideas into products and services, and further magnifying the power of ideas by shortening the time from conception to realization. The work involved in design and implementation of these new systems has resulted in a radical shift of human contribution to work from the physical to the cognitive domain. And this has gradually raised the formal reasoning demands and educational requirements of most jobs. In 1969 Peter Drucker not only described the nature of knowledge work, but also introduced the term “knowledge economy”. In 1999, he elaborated on why and how knowledge work and workers are different from manual work and workers. He also clarified misconceptions that knowledge work is not associated with only people who work in offices and those wearing suits. But rather, a mechanic needs to know how to fix a car or plumbing, a farmer needs to know a lot about climate and crops and animals, and a taxi driver needs to know the city geography well. In fact, today this is referred to as ‘domain knowledge’. Even products which we normally regard as physical objects are mostly made of of knowledge. Kjelle Nordström and Jonas Ridderstråle, in their book ‘Funky Business’, argue that today only 30% of the costs involved in developing a car are accounted for by materials, labor and assembly costs, while 70% of the value of a car is knowledge cost of aspects such as research, development, and knowledge related to processes.
Creation of new tasks
Network and technology businesses like Amazon not only create a small core of high paying jobs, but also initiate a much larger network of lower wage jobs that are not in the core, but in the supply network. The central software of the networked platform is built and maintained by the highly-skilled core. With this, new workers are required at small suppliers who are now able to bring products effectively to markets; new workers are also needed for jobs like package delivery because the customer now has them delivered to the home or office rather than pick these up at a physical store; also, warehouses that no longer handle periodic large shipments to local retailers require new workers to prepare small parcels of same or next day delivery to millions of customers; further, new workers are needed at telecom companies, internet service providers, data centers, energy companies, as also, workers at other suppliers to the internet digital infrastructure. Moreover, the required software also needs a set of workers to constantly update and manage the digital infrastructure in order to make it more effective.
David Autor and Anna Salomons categorize new job types created into the three broad categories: frontier jobs, which, according to them, have close association with new technologies; wealth work that is involved in fulfilling the requirements of the wealthy; and, so called, last mile jobs, which are the tasks that remain manual while majority of the task has been automated. Examples of these remnant manual jobs include delivery services to consumers, picking tasks in warehouses and manually combing social-media posts for unacceptable content. The jobs in the first two of these categories are available mainly to the highly educated, while being decently paid. Only the last-mile jobs, which are competed for by lesser educated workers, offer lower wages and are less satisfying. This is due to their being repetitive and monotonous in nature, and have the potential of being very likely candidates to get automated in the near future.
Automation may also create new jobs which downgrade human work into certain low-skill routine and repetitive tasks to overcome the present limitations of technology. These jobs require humans to complete certain activities that machines are, as yet, not capable of performing. For example, Amazon recently hired 175,000 additional human workers for its highly automated warehouses. These warehouses have robots moving items around the warehouse floor. These workers are needed for the repetitively picking tasks that humans perform easily but where robots have a fundamental issue in their ability to pick up an object placed in front of them. This shortcoming is due to the uncertainty of spatial perception in robots (of understanding where things are in space), and in the uncertainty of control over gripping the object (uncertainty in the grip friction, and the shape and weight distribution of objects). These shortcomings of robots, in turn, get reflected in the difficulty of getting the robots’ grippers to make the correct contacts in space, due to the fact that even a very small error at their fingertip often results in the robots dropping objects. Humans, who are good at image recognition, are also utilized for the routine and repetitive tasks of labeling datasets for machine learning and AI applications. This is because of the limitations of automated image recognition technology devices in distinguishing between, for example, different dog breeds, which can vary greatly in size, color, and shape based on their breed. Humans, on the other hand, with their superior visual perception and recognition abilities, can easily distinguish and classify the datasets of different dog breeds. Hence, humans can be productively used to train the algorithm used in these applications for classification. Once trained, the algorithm can do the job by itself.
Uber and Lyft present an interesting case study of how technology creates new jobs. Earlier, customers perceived the brand of the dispatch company, which may have been sublicensed to many smaller firms, or even a single taxicab, when they saw a branded taxicab. The fragmented taxicab industry provided work not just for drivers, but required a whole range of job categories like managers, dispatchers, maintenance workers, and bookkeepers, who were in secondary support roles for the brand. Uber and Lyft platforms make use of new technology to create efficiency by eliminating the large hierarchy of individual supplier firms and their managers, effectively replacing them with a relatively flat network of drivers managed by GPS, algorithms, network-based reputation systems, and smartphone apps to coordinate driver and passenger. Although, Uber and Lyft provide both the dispatch as well as branding services similar to the taxi companies, they do this more efficiently. However, in comparison, Uber and Lyft essentially subcontract transportation jobs to individuals rather than to smaller businesses, and then charge a percentage of the revenue rather than a rental fee for the use of their brand on the owner’s taxicab. The introduction of technology resulted in a fundamental restructuring of the taxi-limousine industry from one of a network of small firms to a network of individuals, replacing many middlemen in the taxi business with software. This freed up resources to put more drivers on the road. This new technology also increases the supply of workers by making it possible for even individual car owners working part-time, to successfully find passengers and navigate even to out-of-the-way locations not known to them. Hence, following the introduction of Uber and Lyft, the fraction of people who worked as chauffeurs or taxi drivers increased manifold. Uber and Lyft also rely on their network of customers to control the quality of their service through their feedback. Lyft even outsources the talent management function to its network of highly rated drivers for enlisting new drivers. These technologies provide other benefits to society, such as the ability to employ some differently-abled and previously deemed unemployable populations, such as, for example, those hard of hearing, to serve as drivers.
Robotics, artificial intelligence, machine learning, and other technologies provide a lot of promises for increasing productivity, and perhaps, even for the elimination of work that is unsafe, difficult to do, or work that is not very rewarding for humans to take on. This can then free up humans to do more creative work, while leaving the robots to take on some of the repetitive and unsafe tasks. Robotics and other technologies can make roads safer, make the environment more sustainable, and improve human health care. Some futurists, which include Daniel Susskind of the University of Oxford, suggest that artificial intelligence may well eliminate traditional professional tasks, in continuation of the trend established earlier by mechanization and automation technologies.
References:
https://www.asmag.com/mobile/article_detail.aspx?aid=24838
https://www.scribd.com/document/363729895/Whats-the-Future-of-Work
https://wtfeconomy.com/networks-and-the-nature-of-the-firm-28790b6afdcc