Robots, Genes and Bytes: Technology Development and Social Changes Towards The Year 2020

Available online at www.sciencedirect.
com
Technological Forecasting & Social Change 75 (2008) 1176 1201
Robots, genes and bytes: technology development and social changes towards the year 2020
Antonio Lpez Pelez a , Dimitris Kyriakou b,
a
Universidad Nacional de Educacin a Distancia (UNED), Madrid, Spain b IPTS, Joint Research Centre, European Commission, Seville, Spain
Received 31 July 2007; received in revised form 24 January 2008; accepted 25 January 2008
Abstract Scientific and technological policy has become a key activity in contemporary societies. In this context we present different projections about the evolution of science and technology in the area of robotics and advanced automation, which in turn shapes the new possibilities and risks emerging in this area in the future. This goes hand-in-hand with an analysis of the interaction of such trajectories with the social context from which they emanate. This interaction reinforces the need for establishing the probable sequence of technological innovation; analysing the impacts on economy and society; and providing qualified information for decision-making, both in policy and business. In this article, we present the results of the prospective research carried out in the field of robotics and advanced automation, paying special attention to the transformation trends of organizations, and the integration of robots in daily life and leisure, and underscoring potential repercussions which may deserve more attention and further research. 2008 Elsevier Inc. All rights reserved.
1. Introduction We are immersed in a non-stop scientifictechnological revolution, which is changing our world, risking our survival [1], and opening new possibilities in the transformation of our species and post-
The article expresses strictly personal views and not necessarily those of the authors' employers. The authors would like to thank two anonymous referees and UNED's sociology Prof. Jose Felix Tezanos, director of the project 'Tendencias Sociales de Nuestro Tiempo', on which this analysis builds. Any remaining error is the sole responsibility of the authors. Corresponding author. PERSON, IPTS, EDF EXPO, C/ Inca Garcilaso s/n, Isla de la Cartuja, Sevilla 41092, Spain. E-mail addresses: alopez@poli.uned.es (A. Lpez Pelez), dimitris.kyriakou@ec.europa.eu (D. Kyriakou). 0040-1625/$ - see front matter 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.techfore.2008.01.002
A. Lpez Pelez, D. Kyriakou / Technological Forecasting & Social Change 75 (2008) 11761201
1177
human life, genetically modified, in which intelligence, genes and machines are merged, giving rise to what has been called NBIC convergence (Nano-Bio-Info-Cogno) [2]. The prospective analysis about evolution trends in XXI century technology, take as a starting point two fundamental features of present techno science: firstly, the quick growth of innovation processes and application of technologies; secondly, the convergence of technologies that give a higher speed to the technological change, and allow new developments and applications. Robotics is a good example of this process of quick growth, convergence and diffusion of new applications [3]. We can distinguish three orientations that allow us to classify the prospective studies: In the first place, there is research focusing on the internal dynamics of the technological development, and its consequences on human life from a perspective focused on the species identity. Here we can distinguish two different perspectives: first, those approaches that analyse the scientifictechnological development as an exponential process, in line with Moore's Law; second, those approaches that focus on the radical innovation, assuming that, at a certain moment of the scientifictechnological (S/T) development, a new revolutionary era will begin an approach reminiscent of the kuhnian view of paradigmatic shifts in science. In both cases the convergence between biology, robotics and artificial intelligence gives rise to a choice among three possibilities: the genetic improvement of human beings, to the point of improving our biological potential, launching a new age in the species evolution; the development of artificial intelligence in its stronger variety towards eventually producing self-governing machines, robots with the possibility of self improvement and repair, creating beings more intelligent than humans; and a different future in which nano-robots and biological improvements in the human brain will coexist producing a mixture of human being cum robot, bringing in a sense the use of technology by human beings back to its primordial role: beyond a certain evolutionary stage human beings did not adapt biologically, but rather 'adapted' their environment to their needs/goals, by means of their intelligence and technology. By way of example, the Neolithic revolution, particularly in agriculture and stockbreeding, allowed a high increase in productivity of land and animals, and so made possible the growth of the human species [4]. The projected challenges (e.g. climatological) life on earth may face (to a large extent due to anthropogenic activities), and the difficulties in reversing the trends exacerbating these challenges, give extra credibility to the hypothesis that we will adapt not only by changing our environment but also ourselves. The ethical debate about how to face these radical changes in technology and in our own life is present in well-known recent analyses, from approaches that defend, from an optimistic point of view, the benefits for humanity that come from the convergence between genetics, nanotechnology and robotics [5], to those approaches that insist on the impossibility of controlling the technological change and its negative effects e.g. will androids also assume/develop all the negative traits of human beings? [6]. Secondly, there is research that takes as its starting point how societies build their technological trajectories in an interaction process between scientific innovation, citizens' demands, the intervention of relevant actors such as firms or government, and the adaptive and creative response of economical and legal institutions. In this approach, it is of great importance to understand the consequences of existing/ emerging technological trajectories, because once these are established, their inertia limits our possibility of choice. It is also relevant to analyse the paradoxes caused in the innovation, design and application of new technological process, beyond the anthropologically pessimistic or optimistic discourse.
1178
Table 1 Robotics and social change: analysis prospects Research fields Robotic technology Key points
Development of new technologies in key fields for robotics Development of new industrial and services robots with increased abilities, and with an increasingly affordable price, available in the market Automation Percentage of activities that will tend to be automated in the next years by the use of robots Consequences on the organization of activities, so that the robot and/or the automation system can be operational Work organization in the industrial Impacts on the organization of companies and services sectors Impacts on the employment turnover Impacts on security and health working conditions Daily life and free time Impacts of robots as new free time companions Impacts of robots and advanced automation systems in the household and housework Source: Authors' own elaboration.
There is another line of research that along with the analysis of technological possibilities, examines people's attitudes and their interaction with the way in which innovation is ultimately realized/ accepted. By way of example, it was an unexpected demand for 'private', non-intrusive, answer-whenand-as-you-please communication i.e. short text messaging that proved an unexpected source of profitable business for mobile phone operators. In the biotechnologies field, the opinion polls about public perception are very important [7]. Attitudes towards a particular product can render commercially unviable research results in the area of transgenic food. In this type of technology, Peter Singer and Abdallah Daar show how public and private companies and institutions must promote social implication mechanisms [8]; Jasanoff underscores the importance of viewing scientific development and its potential impacts from the perspective of science management [9]. Finally, there is prospective work supported by governments and international institutions whose main aim is to identify the crucial innovation lines in the near future, the competitive position of each country, company or sector. In this approach, technological development is analysed as a basic variable affecting the competitiveness profiles of a country's economic agents, and effective ways to fund/ promote promising research programs [10]. From a perspective whose aim is the balance among these theoretical positions, in this article we present the results of the series of prospective studies that we have carried out on Robotics and Advanced Automation in the European Union.1 We have used the Delphi technique for two reasons. Firstly,
1 For this, we have carried out three consecutive Delphi studies on experts in information and communication technology, robotics and genetic engineering and biotechnologies, in 1996, 2002 and 2005. Fifty experts from each group were chosen every time, what results in a total of 450 experts interviewed using the Delphi method. In each exercise, a previous study, a pilot questionnaire, and two tests of the final questionnaire were carried out. The research on the Robotics area was coordinated by Professor Antonio Lpez Pelez, Ph.D., from the Department of Sociology III (Social Trends) of Universidad Nacional de Educacin a Distancia (UNED), Madrid (Spain). Some of the obtained results have been referred to in the data base of the programme Future Studies of the OECD. In this article, we present the final results of the research exercise on Robotics and Advanced Automation. Previous results were published in [1923].
1179
independently of the debates about the strengths and the weaknesses of the technique [11], it is a well established technique [12], in general widely-used in the scientifictechnological area [13] since the 1970s, and has also been long established in research on social trends [14]. Secondly, leading studies about forthcoming technology carried out by European, Japanese and US research institutions have used that technique [1518]. It must be borne in mind that, after Japan, the five countries with more density of robots for each 10.000 workers are countries belonging to the European Union: Spain, Germany, Italy, Finland and France. We compare these results with prospective research in other geographical areas, analysing the technology evolution trends; the socioeconomic models on the basis of which technology is promoted and taken up; and, specifically, the experts' projection on the main impacts of the expansion of robots and advanced automation systems in the organization of firms, in the transformation of work, in the new requirements of professional education and qualification, and in the safety and health conditions at work. In this direction, we can differentiate between four dimensions in our analysis (Table 1) We must bear in mind that robot and advanced automation are not synonyms, however they are closely linked. The robot is part of a system of advanced automation that intends to recreate human behaviour, developing duties without human intervention, and with a determined level of intelligence and learning. In our research, we analyse the projections/forecasts about the future development of robots, and the ones on those areas of activity where advanced automation strategies are going to be developed, pertaining to the development of new robots. We also address the projections on future impacts due to the reorganization of activities in the industrial sector, in the services sector, and in people's private lives. This is a key question, since the robot needs, in order to be useful, a reconfiguration in the activity involved from the beginning, considering the robot's characteristics [24]. In economic terms, this means that the investment in the purchase and the installation of a robot must consider both the cost of the robot, and the cost incurred through the reorganization of the activity, and its context (according to the International Federation of Robotics, the price of a robot implies only one third of the investment that must be done in software, peripherals and systems engineering). In organizational terms, the installation of robots and advanced automation systems implies a change in the work-management model. Last, but not least, in sociological terms, the incorporation of robots into private life means a reorganization of household spaces and people's lifestyles. Technology is developed and adopted in a symbiotic relationship with cultural, political and economic developments in a given society. Human societies face wide spectrums of possibilities depending on, and in turn shaping, the technological trajectories they pursue. Understanding this symbiotic relationship should inform technology management and related policy choices. 2. Converging technologies and new robots: the future of intelligent machines Technological advances will allow in the next years the design and launch of increasingly powerful robots, which could be able to overcome the present restrictions of this technology (open space navigation, learning capability and intelligent behaviour, integration of sensors and handlers allowing more efficiency for all types of tasks, and appropriate designs to be integrated in a non-industrial environment, allowing a larger role in domestic and leisure services for the population). Converging technologies will allow exceeding the boundary of present limits due to materials, intelligence and machine mobility. Japanese experts anticipate, between the years 2013 and 2027, the development of robots characterized by the use of intelligent systems allowing them to keep and reuse the previously achieved skills and knowledge (Table 2). The human beingmachine relationship will become simpler,
1180
Table 2 Japanese experts' projections on robotics converging technologies Events Development of intelligent systems able to make decisions in the field of human decision-making related to processes, skills, know-how and experiences open to others with the aim of reusing or learning Practical usage of a humanmachine interface in which virtual users with comprehension and dialogue capabilities can carry out tasks/procedures with the informatics equipment through dialogue with humans Development of artificial membranes with similar functions to biological membranes Practical use of the technology used by a computer to control the motor activity of the brain, being able to execute a direct and voluntary control on the artificial limbs without depending on the spine or the peripheral nervous system Development of intelligent robots with the ability to see, hear and perform other sensory functions, able to think, make decisions and act in ways similar to human beings Time horizon 2013 2014 2017 2023
2027
Source: National Institute of Science and Technology Policy, The Seventh Technology Foresight. Future Technology in Japan toward the Year 2030, Science and Technology Foresight Centre, Ministry of Education, Culture, Sports, Science and Technology, Tokyo (Japan), 2001.
and intelligent robots will be developed; those will be able to think, see, hear, make decisions and act in ways similar to human beings around the year 2027. In the next decade, experts project the introduction of robots and advanced automation systems in four large areas (Table 3): first, in traffic control, not only in planes but also in automobiles, increasing passenger safety, and reducing, in the case of air transport, this type of jobs by 50% in relation to the levels in the year 2000; second, the development of research robots able to work under extreme conditions, especially under water, deeper than 10,000 m; third, in the industrial field, implying radical transformations in the employment opportunities and in the working techniques of the manufacturing industries workforce. Finally, in the health and leisure field, microbots can enable minimally invasive
Table 3 Japanese experts' projections on robotics new applications Events Practical usage of systems in which radical automation of air traffic control through informatics and computational technology advances, will imply a reduction of this type of jobs by 50% in relation to 2000 The advances in digitalization and the higher sophistication of industrial robots will imply radical changes in employment opportunities, and in working techniques of the manufacturing industries workforce Practical usage of robots for underwater exploration deeper than 10,000 m Development of minimally invasive surgery techniques, with micro machines or robots, which will be used for most surgeries Practical usage of automatic driving systems able to develop safe and gentle operations in roads, simply choosing the requested destination Common use in Japan of houses with robots and other automated machines, that help old and disabled people with household tasks, including feeding, bathing, going to the toilet, leisure activities, and can, moreover, carry out these activities without human help Common use (one in each home) of domestic robots, able to clean, wash up, etc. Time horizon 2013 2013 2015 2017 2017 2017
2018
Source: National Institute of Science and Technology Policy, The Seventh Technology Foresight. Future Technology in Japan towards the Year 2030, Science and Technology Foresight Center, Ministry of Education, Culture, Sports, Science and Technology, Tokyo (Japan), 2001.
1181
surgery; using in a very common way domestic robots that do the household tasks and at the same time help old and disabled people, can improve their life quality. These results reflect the projections made by the experts interviewed in these exercises. The basic technologies whose convergence allows the design of ever more powerful robots have to do with the capabilities that make human beings more multifaceted: eyesight, intelligent interaction, integration in a natural language system, mobility in open spaces, and the convergence of all these advances, making possible in this way to develop robots with more skills, able to substitute human beings for a higher number of tasks. The advances that will permit the intensive and coordinated usage of technologies related with the capabilities of sight, intelligence, language, and navigation will be produced, according to the experts' projections, in the second decade of the XX century. Note here that intelligent/intelligence or intelligent interaction refers to the capacity to make decisions in relation to executing the tasks at hand, on the basis of environmental information; these robots will have previously been trained for the tasks, and they will process the information received from their environment to reach decisions towards task execution. In 2015, industrial robots with 3D vision are expected to be available in the market (Table 4). In 2020, the present limitations in the fields of intelligent interaction, mobility and integration in a system of natural language are expected to be overcome. And 5 years later, in 2025, robots that will integrate these technologies are expected to be available, permitting a remarkable increase in the capacity for task automation in all activity areas. This projection that places in 2025 the integration of the different technologies coincides with the projection carried out by experts in the Delphi studies made in 1996 and 2002, which points out that it is a
Table 4 Projections on the advances in robotic technology Events Delphi study 1996 Delphi study 2002 Delphi study 2005
Time Security in the Time Security in the Time Security in the horizon projection horizon projection horizon projection Intelligent robots will be available in the industrial sector (what we mean by intelligent is the capacity to make decisions in relation to the tasks done; these robots will have previously been trained for the tasks, and they will process the information they receive from their environment to reach decisions 3D-vision robots will be available Robots integrated in a natural language system will be available Robots able to move around in an environment not previously known will be available Robots with all four characteristics mentioned above will be available 2010 4 2015 4 2020 3
2010 2010 2010 2025
4 3 4 3
2015 2025 2020 2025
3 3 3 3
2015 2020 2020 2025
3 3 3 3
Note: Security in the projection: 5 degrees (1: not certain; 2: slightly certain; 3: certain; 4: fairly certain; 5: very certain).
Fuente: Grupo de Estudio sobre Tendencias Sociales (GETS), Estudio Delphi sobre tendencias cientfico-tecnolgicas, Sistema, Madrid, 1996; GETS, Estudio Delphi sobre tendencias cientfico-tecnolgicas, Sistema, Madrid, 2002; GETS, Estudio Delphi sobre tendencias cientficotecnolgicas, Sistema, Madrid 2005.
1182
comparatively stable projection that is not pushed back in time between Delphi studies. However, we can see that in the case of technological advances related to robots intelligence, a relative deceleration is confirmed: the projection on the effective availability of a robot with that capability is delayed for 10 years, from 2010 (Delphi study carried out in 1996) to 2020 (Delphi study carried out in 2005). The variation in the projections of technological advances, however, does not alter the year 2025 as the horizon by which automation will be able to improve qualitatively, when the technological problems related to vision, intelligence, language and mobility could be solved. This projected horizon, both in the Japanese case and in our research, is reflected in Kurzweil's [5] projections: biotechnology, robotics and nanotechnology will reach maturity around 2020, and around 2050 what he classifies as the four revolutions will give rise to singularity, a new post-human reality. Whether his hypothesis about singularity is accurate or not, is beyond the scope of this paper. What is to be noted here is that the experts agree that the decade of the twenties, XXI century, will be a key moment in the reproduction of the intelligent human conduct, and in the capacity of robots to develop all type of behaviours. 3. Experts' projections on robotics, work transformation, and high automation firms The prospective analysis on the impacts of technology must avoid, from our point of view, both the Scylla of technological determinism and the Charybdis of radical social constructivism applied to technology. Neither is technology divorced from the social context from which it springs, nor is technology a social product that is independent of the viability conditions that science/technology establishes. It is not a fortuitous response to the temporary victory of social strategies that look for a temporary balance (the scientifictechnologic mechanism must, above all, work). Neither does it abide solely to a technological logic divorced from the social reality around it [25]. Technological development and social change are closely related, and the preceding historical sequences, regarding scientific advances, restrict/condition evolution paths in the future. The projections on the diffusion of a certain technology help us deal with one of the structural factors playing a role in the development of our societies. In this light one should see the projection of experts for the next 20 years, emphasising that industrial and services robots will spread increasingly in different activity fields, until becoming a technology widely used in the production of goods and services in the next decades. 3.1. Projections on the diffusion of industrial and service robotics In certain industrial areas, the robot is nowadays a well-known tool that is widely spread. In the automotive sector (Graphic no. 1), the experts project 60% automation of activities in 2010 , and from then on the automation of tasks will grow to reach 80% of all the activities until 2050. If we examine the projections in other activity areas, such as chemicals, oil, coal, rubber and plastic, metallic products, or shoe and textile, in those sectors, too, there are projections of a 70% rate of automation around 2050. In the short term, by 2010, the automation levels will vary between 27.5% of the activities in the area of chemical products, oil, coal, rubber and plastics and 50% of the activities in the field of automation. In the medium term, by 2025, we can remark two characteristics: firstly, the increase in automation of tasks in all fields; secondly, an expansion that tends to homogenize the level of automation reached in the different areas of the industrial activity: automation varies between 65% in all areas of the automotive field, and 42.5% in all areas in the shoes and textiles field. And by 2050, the convergence is even higher: it varies between 60% of all the tasks in the food and drink areas, and 80% in the automotive sector.
1183
Graphic no. 1. Projections on the evolution of the percentage of tasks automated in different industrial sectors. Fuente: GETS, Estudio Delphi sobre tendencias cientfico-tecnolgicas, Sistema, Madrid, 2005.
In this context, a factor that limits the growth of robotics, according to experts' remarks and comments in the Delphi study, is the size of companies: the smaller the companies in the sector are, the longer it will take for automation to be adopted. The horizon for automation, according to the experts' projections, is coherent with the historical expansion of industrial robotics, and with the projections made by the International Federation of Robotics about installation of new robots in the coming years. The automation of tasks will be helped by three favourable factors, according to the experts' projections: the decrease in the cost of the robot, the technological improvement that will allow its usage in more fields and the tendency to automatize a higher percentage of tasks in the industrial sector. In the agricultural sector, experts project an important expansion of the automatic and robotic systems, in an environment that shows specific difficulties: the sowing and harvesting work demand technological advances in mobility, in tactile and visual abilities; the works of classification, canning and storing permit an easier automation, which however can be complex, depending on the agricultural product. In the case of countries like Italy, Spain or Greece, we must point out that agriculture is a sector in which many of the workers are immigrants and with low qualification levels, which implies severe difficulties for those workers displaced by agricultural robots, when looking for a new job. In the short term (Graphic no. 2) experts project 10% automation rates in agricultural activities, which will increase to 25% in 2025 (date when experts project the development and usage of robots more powerful than those at present, with the integration of vision, mobility, language and intelligence). In the long run, by 2050, an automation rate of 35% is projected in all tasks of this sector.
1184
Graphic no. 2. Projections on the percentage of tasks automated in different areas in the agricultural and services field. Fuente: GETS, Estudio Delphi sobre tendencias cientfico-tecnolgicas, Sistema, Madrid 2005.
From the experts' point of view, two trends meet in this process. First, the progressive reduction in the price of industrial and service robots: prices of industrial robots, expressed in constant 1990 US dollars, have fallen from an index 100 to 54 in the period 19902005, without taking into account that robots installed in 2005 had a much higher performance than those installed in 1990. When taking into account quality changes, it was estimated that the index would have fallen to 22. In the same period (19902005), the index of labor compensation in the US business sector increased from 100 to 179. This implies that the relative prices of robots fell from 100 in 1990 to 23 in 2005 without quality adjustment, and to 10 when taking into account quality improvements in robots. Other major robot-using countries had similar developments in their relative robot prices [26]. The progressive reduction of the cost of robots is linked to other structural advantages of the automatic and robotic systems: they can work 24 h a day, they reduce work accidents, they carry out jobs which are dangerous for the workers' health, and they also can help reduce conflicts on labor issues inside companies. These advantages reinforce the interest in the investment on automatic and robotic systems. A second tendency that seems to be feeding the growth in the number of industrial and services robots has to do with the control of the immigration flow. The option of robots is reinforced by the objective difficulties of managing the social and economic integration of the immigrant population. In this light, the conditions to emigrate to the European Union will get predictably harder (recent developments in France, for example, point that way), and the economic strategy that tries to compete by low immigrant labor cost will be hard to sustain in coming decades [27,28].
1185
Construction, one of the important sectors of the economy in the countries of the OECD, is still highly labor-intensive, with workers who, as in the agricultural sector, are often immigrants, both in Europe and in the US. The adoption and diffusion of robots and automatic systems in this field requires addressing first certain technological bottlenecks (e.g. visual capacity, mobility in difficult, changing environments, etc). This is why, as in the aforementioned case of agriculture, it is of special interest to review the projections of the experts. The percentage of activities that will be made by means of automatic and robotic systems is projected to be similar to that in the agricultural sector (Graphic no. 2): in the short term, by 2010, 10% of activities will be automated, and from that moment on, the level of automation will be increased to 20% in 2025, and to 30% in 2050. The replacement of human employment by robotic employment in agriculture and construction, two sectors characterized by the intensive use of human workforce, to a rate between 10% and 30%, if the experts' projections are fulfilled, will have a strong impact on employment in these sectors. This underscores the need to establish educational strategies that permit enhancing the scope of skills of workers in these sectors, to improve their chances of finding new jobs. Service robotics is on an early stage in its development, but it has high potential, and there are already areas, for example medical technology, security or defense, where the prototype stage has given way to a wide offer of automatic and robotic systems generally used to perform very specific tasks. In the service sector we can identify four areas in which strong automation is projected for the coming years (Graphic no. 2): First, the activities related to security, surveillance and defense. In this sector there are strong investments, both in order to protect buildings and key points such as railway stations, airports or nuclear power stations, and to achieve a minor exposure of human beings to fatal risks in war conflicts: automation of military vehicles, soldier-robots, etc. The perception of experts is clear: in this type of activities we will see an increase in investments, and the technological advances will permit, in the short term the automation of 20% of activities by 2010, 40% by 2025, and 60% by 2050. Imagining the security or defense sectors with a level of automation similar to that of an automobile factory shows the enormous projected growth of robots in this area, which entails activities that are both quite visible and of a high symbolic value. Second, the automation of tasks in the health sector. Surgery robots are already used daily in the professional activity of many hospitals. Spanish experts also project a strong growth of robotics in this field: a 10% rate of automation on health related tasks will be reached by 2010, 25% by 2025, and 45% by 2050. Once more, if experts' projections are fulfilled, in the next few years we will experiment an important change in health care systems, ushered in by the generalized use of robots and systems of advanced automation. Third, tourism provides an important percentage of the GDP of the economy in many OECD countries. In this sector, where much emphasis is placed in developing a rapport with the customer, the growth of robotics is an indicator of the technological development projected, in terms of human-like intelligence and mobility. This is quite a step beyond the model of industrial robotics designed to perform specific repetitive tasks in a predefined environment. According to the experts, in this sector a strong process of automation will lead to the diffusion of multifunction robots. In the 2010 horizon, 5% of the activities in this sector will be automatized, 15% by 2025, and 30% by 2050, reaching higher levels of automation than what can be noticed nowadays in a sector such as food and drink, or in the oil industry. In the educational and R&D sector, we will also see, according to the experts' projections, substantial adoption/diffusion of robots. Robots and automatic systems will perform 5% of tasks by 2012, 20% of tasks by 2025, and 30% by 2050.
1186
If these projections materialize, the growth of service robotics will allow the automation of a great number of tasks in the coming years, in sectors that are quite labor-intensive, making the robot a machine more and more integrated in the daily life of people in the XXI century. In this sense, we can ask again the well-known questions about the impact of robotization on employment, and the transformation of the social relations depending on the common presence of robots in different sectors. The emergence of a new companion perfectly adapted to our demands, both at work, as well as in personal/intimate contexts, may mean a structural change in our pattern of life [29]. 3.2. Towards a robotic society? Perspectives on work and organizations in the planning horizon of 2015 In our research, we focus on projected developments towards the year 2015, in relation to three basic questions: employment impacts in a context of strong automation in the industrial and services sectors; the characteristics of company organizations, especially career development, the evolution of salaries, and the evolution of work conflicts; and the characteristics of jobs in the coming years: contract stability, flexibility, the control of work load, and workers' functional mobility, and the evolution of work risks, particularly the psychosocial , linked to the adoption of new technologies - especially robotic and automation systems. 3.2.1. Automation, unemployment and the tertiarization of economy One of the characteristics of economic developments in the next few years, from the experts' perspective, will be the growing incorporation of robots and automatic working systems (Graphic no. 3) at work and in leisure. In this context and from a perspective that links automation with the transformation of organizations and evolving job requirements, companies with a high level of robotization in the industrial service will tend to hire fewer but more skilled employees. As one of the experts pointed out, employment will be smaller but more skilled: companies will demand more qualified personnel each time, and the volume of employment will depend, as ever, on the growth capacity and competitiveness characteristics of each firm. In the best-case scenario, the production of goods and the service capability will increase with the same amount of personnel. Although other posts related to automation processes will be created, there are two issues/caveats raised by experts: first, those employees whose post is suppressed by the introduction of robots, may have to adapt and perform in quite different types of jobs; and second, the characteristics of the alternative job offer can imply a loss in job security, worse working conditions and a decrease in the salary. However, in the services sector the experts do not project a fall in employment by the year 2015. From their point of view, what will happen is that there will be an investment in all activities open to automation, but new services will appear and these will compensate the theoretical decrease in employment due to automation. This is a projection that has been maintained/reaffirmed in the last two Delphi studies. What is expected is the replacement of human staff by robots in certain jobs - a phenomenon that they project will be remarkably increased in the next few years. This will be compensated by strong growth and overall employment creation in the services sector - a positive 'income effect' as it were, that will counteract the negative 'substitution effect' of the diffusion of robotics. 3.2.2. Changing trends in the business organizations The projected pattern of firm organization over the coming years in research about the transformation of business organizations in technologically advanced societies underscores the importance of flexibility.
1187
Graphic no. 3. Projections about the percentage of activities carried out by robots and automatic working systems in 2015. Fuente: GETS, Estudio Delphi sobre tendencias cientfico-tecnolgicas, Sistema, Madrid, 2005.
Flexibility becomes an essential characteristic throughout the firm. As the Third European Survey on working conditions shows, flexibility applies to all working aspects: working hours (non-stop work and part time jobs), work organization (multi-faceted, team work and capability) and type of employment (18% of all paid employees in 2000 had temporary contracts) [30]. Firms are projected to be organized around less hierarchical structures, with a stable nucleus of employees and executives, of strategic importance to the organization, and a series of concentric circles around this nucleus, where working conditions vary substantially. According to the experts interviewed, by the year 2015 the possibility of a traditional career will be uncommon, while, on the other hand, other aspects (salaries, instability, stress) will be similar as in the present. In a context of growing deregulation, contractual instability is accepted by workers in the European Union as an inevitable characteristic of the labor market, and the stress level does not vary with the integration of new technologies. 3.2.3. Advanced automation and working conditions Automation does not uniquely determine but does influence the evolution of working conditions. However, the pattern created in countries in which there has been an intense process of robotization in the last 10 years, is characterized by the worsening of working conditions [31]. Particularly, the evolution of working conditions in the European Union presents the following characteristics (Table 5): higher presence of new technologies, higher pressure on the job (growing in all EU member-countries), higher autonomy at work, and an increase in the number of high-stress jobs.
1188
Table 5 Main results on working conditions in the European Union - There is a direct relationship between health problems and bad working conditions, particularly coming from a very intense/ stressful and repetitive work - The exposure to risky physical factors (noise, vibrations, dangerous materials, heat, cold, etc.) and work in badly designed posts (manipulation of heavy loads and difficult positions) continue being frequent - The work is more and more intense each time: more than 50% of workers make their tasks at a high speed, and in a very accurate period for at least 25% of their working time - The amount of people working with computers has increased: from 39% in 1995, to 41% in 2000 - Flexibility is quite spread in all working aspects: working hours (non-stop work and part time jobs), work organization (multifaceted, team work and capability) and situation of employment (18% of all paid employees have temporary contracts) -Temporary workers (paid workers with contracts of different length and hired by temporary work companies) continue mentioning a higher exposure to risks than those with stable jobs Source: Foundation for the Improvement of Living and working conditions, Third European Survey on working conditions, Dublin 2001.
In this context, the experts project that the introduction of automatic and robotic systems will not modify present working conditions (Table 6). The improvements introduced by robotics (higher quality, increase in productivity, fulfilment of tasks which are hard and dangerous for humans) do not imply an improvement in working conditions for those continuing in their jobs. Instability will continue as at present, as will intensity/stress levels for workers and similarly for salaries, etc. In agreement with this perception, the experts project that the number of workers suffering from stress-related conditions remains the same as at present. They only project an increase in the functional mobility of the workers who use robots (where functional mobility refers to the increase in the number of tasks, changes in job assignments, implying higher qualifications, and a multifunctional worker). This implies that automation will favour a more multifaceted and higher-skilled profile, although working conditions will not necessarily improve in tune with higher qualifications. On the subject of work-related exposure to risk, the experts project a reduction in the number of work-related accidents. It is one of the main objectives that automation can achieve. 4. A prospective look at robotics: resilient old issues and surprising new ones The trends projected for robotics and advanced automation analysed in this paper set the stage for outlining possible repercussions which may deserve more attention and further research. These include techno-economic aspects which may upon closer examination point to surprising possibilities we will present only a few, spanning the spectrum from more technically oriented to more socio-economic-impact oriented ones. In what follows we will use standard economic analysis tools, which have well-known strengths and limitations. The long experience that analysts have in their use provides a common platform and facilitates comparisons. As long as analysts keep in mind the aforementioned limitations and the authors do such tools can be very useful in revealing key dimensions of issues arising due to techno-economic developments, and the very real ways in which they shape society, and the choices/tradeoffs regarding the sort of society which may emerge.
A. Lpez Pelez, D. Kyriakou / Technological Forecasting & Social Change 75 (2008) 11761201 Table 6 Projections on working conditions by 2015: trends projected by experts Trends
1189
Delphi study 2005 Trends Security in the projection
In the next 10 years will the task-attention level required of the workers who use robots be higher, lower or the same as today? In the next 10 years will the workers who use robots have a higher, lower or the same degree of control on the task done? (control refers to the autonomy to decide the rate, the sequence and the means to make the task) In the next 10 years, will the functional mobility of the workers who use robots be higher, lower or the same as now? (Functional mobility refers to the increase in the number of tasks, and changes in the post, that imply higher qualification, professional polyvalence and a multifunctional worker) In the next 10 years, will the level of saturation of the workers who use robots be higher, lower or the same as now? (Saturation refers to the higher psychical effort coming from the multiplication of tasks, higher variation, the assumption of more responsibility, the increase in job uncertainty and longer, more flexible and irregular working hours) In the next 10 years in the companies with important levels of automation, will there be a higher, lower or equal amount of work-related accidents as today? In the next 10 years, will the number of workers who use robots and suffer from stress be higher, lower or the same as today? Source: GETS, Estudio Delphi sobre tendencias cientfico-tecnolgicas, Sistema, Madrid, 2005.
Same Same
4 4
Higher 4
Same
Lower 4 Same 4
What does the future hold for robotics? What is the next step, or the next technological boundary to overcome? The general trend for computers seems to be faster processing speed, greater memory capacity and so on. One would assume that the robots of the future would become closer and closer to the decisionmaking ability of humans and also more independent. Experts disagree as they have since artificial intelligence emerged as a field on how soon we can expect to have conversations with androids and have them do all our housework. Initial enthusiastic optimism in the sixties was toned down after the infamous EnglishRussianEnglish translation experiment. This issue is still not solved (though important progress has been made since then) as anyone who has used machine translation can confirm. Another difficult design aspect, and another incongruence and perhaps a less well known source of frustration about android robots is their ability to walk around on two legs like humans. A robot with biped movement is much more difficult to build than a robot with, say, wheels to move around with. The reason for this is that walking takes so much balance. When you lift your leg to take a step you instinctively shift your weight to the other side by just the right amount and are constantly alternating your center of gravity to compensate for the varying degrees of leg support. Presumably, once robots have the ability perform a much wider array of tasks, and voice recognition software improves such that computers can interpret complicated sentences in varying forms (in terms of dialect, diction, grammatical, accent, etc.) we will see robots as our broad-spectrum assistants (as opposed to narrow spectrum tasks for industrial robots). A stocktaking moment emerges here: not only neither of these two key challenges human-like intelligent info processing (software) and biped movement (mechanical) has been fully tackled in 40+ years, moreover, the mechanical one that would initially have seemed less daunting (since it was cast after all as a Newtonian Physics problem) is proving quite difficult. As researchers like MIT's
1190
Rodney Brooks [32] have argued, interacting with the physical world is at least as difficult as symbolically reasoning about it. This raises the next and larger stocktaking question: on the one hand it is heartening that robots mimicked an elementary homo habilis, using any tools available in a room in ingenious ways to achieve their goal. On the other hand however, does it have to be the case that robots in their evolution/ development follow the human evolutionary path? This question would apply equally to efforts to mimic the human brain, or skeletal balance, etc. Indeed, in many ways this line of thought prevailed in artificial intelligence and robotics in the seventies and led to an emphasis on large, and arm-like industrial robots used on factory floors, and on expert systems focused on dealing with very specific types of queries, but no pretence to emulating human intelligence overall. Advances in software engineering and even more spectacularly in processing power, accompanied by reduction in cost and size, in the eighties and nineties made tough artificial intelligence (AI) and android-design questions popular again. Expert AI systems are still in use especially in areas where traditionally memorisation was a key part of human professional expertise e.g. law and medicine. Their limitations however emerge precisely where the question calls for combining data and theories from different areas, for deviating for standard rule-of-thumb processing, and for willingness to pursue/recognise serendipitous discovery. A key recurrent underlying theme emerges here: Is there any way we can program in robots the bases for serendipitous discovery? Bernard Saw is often said to have quipped that it took him a good 3 weeks to write a spontaneous, impromptu speech, but aphorisms aside, isn't planning for spontaneity, programming creativity a contradiction in terms? A general projection that can be underscored here is the emergence of cognitive science or sciences as the new interdisciplinary queen of the sciences, involving neuroscience, philosophy, linguistics, psychology, computer science, mathematics, molecular biology, etc. As all interdisciplinary approaches it suffers from two well-known problems: first, each discipline tends to quickly create its own jargon to delineate itself from others, fostering a sense of tribe among its practitioners, and making its theories (and their seriousness) impenetrable outside scrutiny; second each discipline tends to treat others with an air of superiority, with its practitioners falling back into their comfortable fold when times are rough. Nevertheless, isn't employing such an interdisciplinary approach necessary? To give but an example, one of the most important developments in neurobiology in the last few years has been the identification in primates of mirror neurons, neurons in the frontal lobes of the human brain that fire not only when the subject engages in a certain activity but also when the subject perceives others engaging in the same activity. It is projected that this is a precursor of similar human brain mechanisms, at the root of human linguistic abilities, the general ability to extract patterns, and intent. Leaving these more general questions aside let us turn to more specific thought-provoking observations/ projections about robotics. When hearing the term robots most people think of quite large industrial arms or human-size androids, nano-robots or nanobots may hold even more promise however, especially in terms of medical applications (think of them travelling inside the blood stream to deliver medication exactly where needed, when needed, and or attack clogged blood vessels, etc.) As mentioned earlier, about traditional humanoid robots: getting them to walk on two legs with the fluidity and balance of a human being is very tough. If walking is the problem for humanoids, swimming may be a more relevant challenge for nano-robots. When things get really small, they don't move through fluids with quite the same ease. Nano-robots may be able to locally deliver cancer-fighting drugs directly to tumors or clot-busting machinery to the site of a blockage. To reach those ends, engineers need a way of moving tiny cargos
1191
through fluids a task that turns out to be a bit more challenging when filmmakers yield the territory to the mathematical modellers. When objects have vanishingly small masses, the effects of viscosity become far more important than the effects of inertia. The upshot is that in the nanoworld, there is no such thing as glide. Moving even a micrometer-sized object through water becomes a lot like trying to breaststroke through honey. Move down the scale to the nanometer realm, and the problem is even worse. In order to move a nano-swimmer needs a nonreciprocal motion, something in which the movements are never symmetrically reversed the problem is often solved by using a flagellum, or whip-like drive, but the mathematics and the molecular engineering of such a system are daunting. Looking at how earthworms move through soil can provide simpler ways to model the movement of nano-swimmers. This may be mathematically simpler and the nano-structures involved may be more straightforward to construct, but they would still need to be connected by some kind of active spring-like molecule and that will take delicate engineering. Then there is the problem of Brownian motion. For truly nanoscale machines to function in a fluid, they would need to overcome not only the stickiness of this highly viscous environment, but also the continuous buffeting of molecules close to their own size. So even if the equations say such a device could travel in a straight line, its progress is likely to look a lot more like an infant on a crowded playground. 4.1. Humanrobot interaction: fading boundaries? The director of the MIT Artificial Intelligence Lab, Rodney Brooks suggested in 2002 [32] that future robotics applications may include: implanting of an identity-confirming computer chip under your skin, replacing national IDs, credit cards, and passports; enhancing vision with retinal chip implants (night vision for soldiers, infrared vision for thermal sensing for firefighters; plugging your brain directly into the Internet, through mental service providers, descendents of ISPs, Internet service providers; replacing organs, first external, then internal, from in vitro lab growth; genetically-engineering miniature robotic sensors in E. Coli cells for blood stream patrol duty (sensors for pH, light, electricity, magnetism). Others [33] examine the effects of wireless linkage, global interconnection, miniaturization, and portability on our bodies, our clothing, our architecture, our cities, and our uses of space and time. This brings us to a third projection (though in this case retrojection could perhaps be coined): Computer viruses raise all kinds of interesting issues. When the term was borrowed from medicine in the mid-eighties to describe the program propagation among computers, intended by the virus-creator but not the carrier users, some thought it was too strong a term to use from too distant a discipline. The term caught on, and if implantable chips can catch computer viruses, which can in turn cause medical lesions to their human carriers, the term will have come full circle: a medical term borrowed by computer science will acquire through computer applications a new medical aspect. This is a projection which brings its subject retrospectively back to where it initially belonged. Humans can fully incorporate tools and supporting cultural practices into their existence. Technology as simple as writing on a sketchpad, as familiar as Google or a cellular phone, and as potentially revolutionary as mind-extending neural implants all exploit our brains' astonishingly plastic nature (the measurement of time, for example, has wrought enormous changes in human existence), as well as ways in which increasingly fluid technologies can adapt to individual users during normal use. As we enter an age of wearable computers, sensory augmentation, wireless devices, intelligent environments, thought-controlled prosthetics, and rapid-fire information search and retrieval, the line between user and tools grows thinner day by day. The recent reviews/discussion of Kurzweil's 2005 book The Singularity
1192
Is Near: When Humans Transcend Biology, in the pages of the Technological Forecasting and Social Change [34] provide a thorough discussion, for those further interested in the matter. A world governed less and less by boundaries and more and more by connections requires us to reimagine and reconstruct our environment and to reconsider the ethical foundations of techno-economic decisions straddling such boundaries, and the way in which laws such as privacy laws can be adapted/ observed in such a setting clearly national cultural traditions are important here, regarding attitudes towards technology but also towards the legal framework and its observance (France, Germany, USA, etc). Besides these prospective impacts of robotic developments (including privacy, health, political, etc. aspects), there is an even more subtle one that we would like to explore next. It involves intelligent agents and the impact their future extended use may have on access to goods, services, and to the way our societies negotiate their tight embrace with market transactions. 4.2. Abstract robots: intelligent agents Our fourth projection on impact, regards intelligent agent robots. Here the kind of robotic developments on which we focus are the software, intelligent-agent type, not necessarily accompanying the more standard industrial-arm model of robot, which has been more often discussed and analysed. Robots of the intelligent agent variety, serving as gatekeepers, have been often touted as the killer application in the context of the emerging information society. However, the very emphasis on a single killer application may be misplaced. Consumers may want extensible packages more than a single application. Receiving information from the Internet is like quenching your thirst by drinking from a fire hose. Information overkill may bring back to the fore not only filterers (and thus create job opportunities) but also make us realize the crucial distinction between information barrage and knowledge acquisition and creation. The western conception of knowledge places more emphasis on explicit, identifiable, storable and retrievable knowledge. It underemphasises tacit forms of knowledge which can be deeply ingrained and hard to share. Knowledge incorporates the hunches, skills and insights of the workforce, and not merely the mass of undifferentiated data, threatening to bombard us into fact-filled oblivion. As the adage goes facts can be like the lamp-posts for the drunkard: they are better for support rather than illumination. The socioeconomic impact of intelligent agents which we will explore here is related to the issue of pricing. Pricing is an excruciatingly difficult issue in the information society (IS), with not only business but also social repercussions. At a more fundamental level however, the core difficulties lie not only in pricing connections, but also information itself (and as we have already indicated the two are increasingly blurring into each other in the IS). There are two kinds of costs in producing information: first the cost of creating the first copy, second the cost of dissemination. The key issue revolves around a time inconsistency problem. Namely whereas the producer would like each copy sold to reflect both kinds of costs and thus recoup them fully (and make a profit), for each consumer, and for society as a whole, once the first-copy costs have been incurred, it is non-optimal to charge more than the dissemination cost per copy to each consumer. The reason is that whereas information sharing and idea circulation has positive externalities for society as a whole, by incorporating some part of the first copy creation cost into the price, consumers who can not afford it or do not value it that high will be turned away, thus reducing the overall positive effect to society. The time inconsistency lies precisely in the fact that this analysis is valid after the first copy has been created and its costs have been incurred. If the producer however knows a priori that he can not recoup his
1193
investment then he will not embark on this process and there will be no information to disseminate. Put differently this statement restates Arrow's classic result suggesting that imperfect appropriability of the fruits of R&D leads to its underprovision (provision below a socially desirable and achievable level). What is locally efficient after the creation of the first copy is not efficient when the whole period and the whole process is considered. The Internet presents pricing problems in an extreme form. By being owned by everybody and nobody at the same time, it resembles the commons (air, oceans, etc.) and may even share some of their misfortune as public goods, readily enjoyed by users who however are quick to freeride when they need to contribute to its upkeep either through positive action or through refraining from carrying out harmful activities. Imagine a good that is being sought after by many users at the same time this could be bandwidth for a transaction, or tickets to be purchased, or valuable information, etc. A way to go about this would be to use intelligent agent-robots and priority levels (priority tags) purchased by the user/subscriber either a priori for a period of time, or on the spot. Indeed multiple priority switches have been built by telecoms, allowing multiple tariffing and customised service grades. Higher priority packets would flow through the mail without delay, whereas lower priority ones would have to wait in line. Actually the situation is a little bit more complicated because it is conceivable that many high priority packets could arrive at the same time. Then if traffic congestion calls for it, there would have to be a way to prioritize among the packets with the same high priority. One way would be to use an auction mechanism to grant highest priorities to the highest bidders (and/or their representative intelligent agents/ robots, presumably those who value speed more, and, crucially, who have the wherewithal to afford raising the quantity bid. Auctions could solve the problem only to give rise to another one: a bank can probably afford to pay more to raise its priority rating during such auctions, than an individual whose doctor is receiving information and guidance during an operation. This is a good example of how analysis rooted in mainstream economic thinking can reveal important cases in which market mechanisms can produce clearly socially undesirable results, and socioeconomic tensions that need to be anticipated/addressed. It also gives an indication of the meaning of a meaty theme underlying this analysis: what society one pursues, e.g. regarding prioritization and access to critical infrastructure and its being subjected to the cash nexus, as in the example above. Obviously this escapes the confines of the Internet and ushers us into the deepest thorniest aspects of pricing and prioritizing in the IS. Depending on the narrowness of the available bandwidth any network or subnetwork of the IS infrastructure may be faced with congestion and prioritization problems. A solution such as penalising the more bandwidth-demanding packets by putting them at the end of the line will effectively abolish one of the most attractive applications, namely tele-medicine, since the latter requires the fast transmission of bandwidth-hungry radiological images. In other words new issues will arise regarding prioritization and order of processing at times of congestion (i.e. intelligent agent-robots acting as traffic controllers may be needed). Traditional methods (First-in-first-out, Round Robin, etc.) may have serious ramifications when speed is of the essence for competing vehicles on the highway. If a prioritization reflecting a regulator's or the government's preferences is put in place, preventing the purchasing of highest priority by those who value it and can afford it, then it is quite likely that wealthy private agents (e.g. financial firms) will partly opt out of this arrangement, and stay with more expensive, and possibly speedier networks. There is nothing wrong with that except that it will lead to cream-skimmimg, leaving the more egalitarian, possibly publicly run network with the least-attractive customers.
1194
Furthermore it would be hard for a government to impose each priority hierarchy on all networks, because information networks of the type envisaged here are rife with cross-jurisdictional uncertainties. Legislating on Internet related issues is an indication of how difficult the process is. The issue of prioritization and auctions points to another repercussion of the IS. The IS will bring down barriers to achieving the economic outcome where the price is bid up to levels where demand meets supply (market-clearing levels). In this sense it would provide a strong boost to competition and efficiency, by facilitating information about demand, supply and prices to reach the interested parties, and by minimising extraneous, waiting-in-line costs. (It actually may bring us close to the Walrasian auctioneer model of the world.) The problem is that such a development may subject to the cash nexus more than society is willing to accept, as the earlier example on tele-medicine indicated in dramatic fashion. To provide a less dramatic example, imagine that users have the ability to run applications which will act as their robot stand-ins, waiting in line at virtual ticket counters (this is neither as far-fetched nor as trivial as it may sound at first it is related to an example used by Microsoft's chairman Bill Gates recently). As soon as tickets become available they are supposed to immediately buy for their masters. How will conflict be resolved if there are 1000 such stand-ins and only 100 tickets? Auctions will do the trick; they will also however essentially prohibit poorer users from ever getting to see any popular performance, as the price will rise to bring equilibrium between supply and demand. In general, monetization through auctions will wreak havoc in those cases where society has opted to fix prices at non-market-clearing levels, and to allow rationing (and determination to wait in line) to filter those who will from those who will not enjoy a service. Several reactions to this development may be imagined if society does not acquiesce in such monetization (including more primitive solutions, such as abolishing advance purchase). Here again what is highlighted is key social issues/choices, regarding the sort of society which may be emerging. The pursuit of efficiency in the use of intelligent agents can exclude from certain activities people who, in a different technological context, can pay by offering some of their time, instead of cash. In many ways societies still preserve niches/enclaves in which it is still not possible to obtain everything through outbidding others. As these niches disappear, more human activities become subjected to the cash nexus, and the thornier the social issues that will be raised. 4.3. Employment Possibly the strongest concern with respect to the impact of the MIS, as well as any technologically driven innovation/transformation, are concerns about employment impact. The topic of Technology's impact on Employment (TE) never fails to generate interest if not outright animosity; in terms of longlasting fascination it ranks together with Malthusian limits to growth, as a topic that will simply refuse to die. At least since the beginning of the industrial revolution people have expressed fears that machines would abolish jobs. The fears became increasingly more pronounced as suffrage was extended to enfranchise increasingly large parts of the population. In the early 19th century Fulton's steam boat attracted the wrath of ferrymen, and Luddites treated jennies and looms as sworn enemies. In the 1940s computer pioneer Norbert Wiener forecast that computers would bring about a depression worse than the Great one of the 1930s. More generally, and less pessimistically, part of our employment problems now come from the fact that technical progress has been unskilled-labour saving to an extent that was not matched by the progress of education. The choices we have made and are making in education par excellence a sector where the
1195
role of government is important condition the employment impact of techno-economic developments, and bring us back, adding even more substance to the recurrent theme of the kind of society one desires. Education has been for a long time one of the main avenues for escaping the predicament of the socioeconomic stratum in which one was born; depending on the way it is structured and adapted to technical change it may instead consolidate and perpetuate cleavages. Beyond the skilledunskilled dichotomy, there is also the short termlong term juxtaposition. In the short term technical progress may destroy jobs but the increase in productivity (where by productivity we mean total factor productivity) and in disposable income leads to increases in effective demand, and eventually to the creation of new jobs. Nevertheless it is clear that adjustment and transition cannot be painless. The creation of jobs may be faster when growth is faster, but when technical progress is responsible for growth, it may also be accompanied by job destruction through skill obsolescence, and the bankruptcies associated with the process of creative destruction. In any case technical progress and productivity growth do not have to be associated with high unemployment; in fact the high productivity growth post war years (what the French call les trente glorieuses) were accompanied by very low unemployment. The response of technology pessimists is that the new worker-displacing technologies informatics, telecoms, etc. are of a new employment-devouring variety. Again, the issue of the sort of society pursued emerges substantively: the way in which the benefits from increased productivity are reflected in incomes across socioeconomic strata is an important factor shaping society. For the last 200 years technical progress has been fast. At the same time incomes in countries leading technical change have risen across the board, the workweek has been drastically reduced (think of 40 hour-weeks, overtime pay, weekends off and holidays) and remunerated adult employment has risen (think of the incorporation of women in the labour force). This has not been achieved automatically: the benefits from increased productivity due to technical change have been arduously fought over, shaping society in the process. 4.3.1. On the one hand, the good news is The situation is not as hopeless as techno-pessimists make it seem. Regarding the speed of introduction argument, although the US has been adopting IT faster than Europe, the unemployment differential between the US and Europe has not reflected this speed differential. And the allegation that most new US jobs are McJobs (of the low-wage, no-future, hamburger-flipping variety) is not necessarily supported by data. Most importantly however, the implicit assumption behind such pessimistic claims is that there is a fixed amount of output to be produced. What is not taken into account is that technological progress generates new wealth; increased wealth leads to higher effective demand, causing increased investment and labor hiring in order to satisfy this increased demand (most typically through selling new products or services). Technical progress that leads to product innovation therefore is beneficial to employment. It is slightly more complicated with process innovation. Process innovation renders the combination of the factors of production more efficient, increasing factor productivity. As a result one or more of these things will happen: either the price of the good or service will fall; or wages will rise; or profits will increase. All three lead to a rise in consumer purchasing power and an increase in effective demand, engendering new investment and labor hiring to meet this extra demand. Often the new jobs will be in new, totally unforeseen industries. The need to train people to use machines and keep updating their skills will generate quite a few teacher/trainer jobs. The entertainment/information services sector will receive a boost through the IT revolution itself (especially the creative side of it, which is irreducible to repetitive tasks).
1196
Moreover, there will be possibilities to having humans and robots work together in new, original ways, emphasising complementarity in their abilities, tireless precision for the robots, and uncodifiable, tacit knowledge on the part of humans. For instance, agricultural workers can precede robots, marking those items ready for harvest, on the basis of their accumulated uncodifiable experience. 2 The costs in terms of lost jobs are more palpable because they are localised sectorally and even geographically at times; the benefits on the other hand are more widely spread out exactly because of the new product and higher demand process explained earlier. The technological boom has been rather beneficial in terms of job creation, as the experience of the fastest new high tech adopters indicates, most notably the US, where growth in productivity, in GDP, and in jobs since 1995 is attributed by econometric studies to a large extent to investment in ICT, and in productivity growth, very notably in the ICT industry itself, as well as in ICT-using industries. The occasional failures of large investments in IT to boost productivity and thus wealth creation, may be due both to antiquated management approaches as well as to our failure to measure qualitative improvements in services due to IT, as opposed to quantitative ones. 4.3.2. On the other hand however Though in the long run technology will give rise to more jobs than it destroys, in the short run certainly no one can argue that every displaced telephone operator will be available to work as a computer repairman. The transition and adjustment period will be rife with painful mismatches. The way in which society, through its policymaking institutions, decides to address such transitions/adjustments will be crucial in cushioning them. Overall, new technology leads to income redistribution. Innovation that reduces the unit cost of production may render certain skills obsolete, and will require a different set of skills. Although the overall change has beneficial effects for society as a whole, since it increases wealth building opportunities, if laborers can not easily change type of skills, new technology will bid up wages for one type of worker and depress them for another, for whom it may also lead to unemployment. Education doubtless is key for the achievement of smooth transition, for staking out the new job opportunities and exploiting them. The more controversial issue is the type of education best suited to the changing environment. The standard apprentice schemes, teaching or retraining people in order to endow them with a particular skill, may make them more vulnerable if this particular skill is of a mechanizable, repetitive nature. Blue collar crafts, that incorporate a certain creative, tacit-knowledge component, may be less mechanizable and hence more resilient with respect to automation and technological advances (cf. example above on agricultural workers). It is important to underscore here that this does not apply only to the highest-skilled, most intellectually gifted jobholders. Rather it applies in cases white collar or blue collar in which creativity (unmechanizable, uncodifiable) emerges in at least part of the job. And such jobs can be white collar (e.g. teachers) but also blue collar (e.g. the agricultural workers mentioned above). Low-level and mid-level white collar tasks have suffered in recent restructuring periods, largely because many of their skills were found to be mechanizable or redundant. The key to avoiding exposure to the arrows of such fortune is to provide value-added in a creative, non-mechanizable way. This is indeed, presumably, the case as we move towards the high end of white collar tasks, where strategic thinking, and creative intellectually driven endeavors dominate. It is still human thinking that comes up with a breakthrough or a creative solution to a vexing problem, and not laptops or wireless modems. As long as artificial intelligence
2
We thank one of the anonymous referees for this example.
1197
does not solve the creativity riddle (and we are a long way from such solutions) the most resilient shield from technology-driven skill obsolescence is creativity and the ability to apply it in multiple contexts. And the best way to promote this is to guarantee a strong education, enhancing technological literacy and numeracy to the extent that workers are literate and numerate enough to be able to be trained and retrained on the job, and to have the basic skills that will permit them to learn new skills even at a later age. 4.4. The gloomy scenario The enhancement of processing power drives the creation of robots that will take over many jobs. The unemployment rate then skyrockets as cheap robots push expensive humans out of half the jobs that we see in our economy today. The automated checkout lines and kiosks that are popping up in places like Home Depot and McDonald's are the first messengers of this. First to go will be expensive, easily robotized jobs in cut-throat competitive sectors, where even small cost-cutting advantages count, and are quickly emulated. This may well include millions of service sector jobs (and not just manufacturing with which robots have been usually associated). The standard retort suggests that robots will create more jobs. Backhoes and bulldozers replaced all the people who used to dig ditches, but then those ditch diggers got jobs building backhoes and were a lot better off. However, the pessimistic scenario notes, in this case, many of the new jobs created may be jobs for robots rather than for humans. 4.5. Towards escaping the gloomy scenario Increased efficiency because of technical advance leads to higher overall income; whether it will lead to a reduction in jobs/income for some people depends on the kind of society people choose to build. Shorter work weeks and higher salaries may very well be the happy outcome of extensive use of robots; however it will not happen unless there are strong economicpolitical forces pressing for it. It should be kept in mind that pay raises or reductions in the workweek do not follow automatically from technical progress, especially if millions of people are unemployed as we saw in times of technical progress during the 18th19th centuries not leading automatically to workweek reductions. Again the underlying question is the kind of society envisaged. The true non-robot-related nature of the problem appears: the politicaleconomic context allowing technology to free us from the obligation to devote a large part of our lives to tasks most of us would not choose to do, if it were not for the money to be earned. Shorter and better paid work weeks would be a natural outcome of technical progress, but it has rarely proved automatic in history. Those who can, usually postpone sharing the windfall gains from technical progress, which accrue to them. Here moreover another more recent and more deep-seated development of the last few decades has become relevant: at some point early in the second half of the twentieth century in the US, and probably later in Europe the trade-off between extra income and extra leisure began to be seen differently. The marginal rate of substitution between leisure and income changed i.e. at their level of income/leisure people began to value an extra unit of income more than an extra unit of leisure, compared to how they used to value them in the past. There is anecdotal evidence on this involving US union leaders discussing overtime work in the late fifties, and calling overtime a terrible practice, of which however their union members wanted their fair share. Similarly, in Germany in the nineties, to avoid/reduce layoffs unions agreed to reduced workdays (and reduced income) in car manufacturing. This, in towns with large populations of car manufacturing
1198
employees, led to new competition for the town's electricians, plumbers etc. by those car manufacturing employees with relevant skills, and suddenly extra leisure time on their hands. In closing, robots can impact not only privacy, access, etc. as explained in the first part but also mechanizable, computer-reducible jobs, not just blue-collar, but also increasingly white-collar ones. Indeed blue collar craftsmen, whose dexterity robots would be hard pressed to emulate, and/or their creativity they could not match, will be less threatened than white collar workers performing repetitive, easily codifiable, computer-reducible tasks. As mentioned above education and creativity-employing jobs will be the key defense to this. On the other hand, robots hold the promise of deliverance from the need/obligation to work, provided increased total income is channelled to reduced work week and/or higher income for all. Citizens can devote more time to being that, in a way not seen since the days of Athenian democracy (when slaves played the role robots may play in the future). Again the underlying question is what kind of society is pursued through democratic politics, pursuing ways to distribute benefits brought by technical change. This process and the period during which it will take place will be marked among other things by the availability of non-mechanizable types of work for human beings, and of creativity-boosting technologyupdated education. A key issue would be addressing the potential exacerbation of income gaps and the subjection of an ever larger part of social interaction to monetary processes. Paraphrasing Niebuhr's inspiring words on democracy: it is technical progress that makes exploring alternative futures possible; it is the way technology may be used that makes such exploration necessary. 5. Conclusions The highly probable affirmations about the future, that create a structure or system, allow us to establish the characteristic features of the future society. In the creation of scenarios, it is about searching for the convergence of evidence and highly probable events, but there is not a technological determinism that implies an unavoidable historical sequence. With regard to the probable technological evolution, the social processes of adaptation, recreation and modification of the technologies will determine the real characteristics of the future society. That is why, the planning horizon of the technological projection and its social impacts show us a sequence that offers information, and that signals social problems that we will probably have to face in a not-too-distant future. In this way, it allows us to collect critical information to be able to act again as democratic societies, and open the possible space when facing the consequences of what is more probable. The creation of scenarios, addresses, for this reason, one of the goals of projection in social sciences: it makes possible to show available options and their conditions and consequences [35]. From this point of view, with the information given by experts, we can build a trend scenario, which we will probably face in the future. The trends scenario derived from the results of our prospective study is consistent with the projection made from other institutions, such as the National Institute of Science and Technology Policy of Japan, and it places us before a historical sequence with the following features: In the next 10 years, the growth of robotics will continue strongly in industry, and its presence will grow dramatically in the services sector. In different areas of activity, from agriculture to construction, the automation of tasks will reach an important percentage, not less than 10 of the total number of activities made in each sector. The technological innovations will be integrated in an economic model
1199
characterized by what has been called permanent instability, which will incorporate without any problem the potential of robotics. On the other hand, a reorganization that would permit a more equitable redistribution of the wealth generated by a more productive economy cannot be assumed automatically as for past technical progress, distribution of benefits will be fought over in social/ political arenas. By 2025, it is projected that the use of robots can integrate technologies allowing the automation of a higher number of tasks. The convergence of advances in artificial intelligence, language, mobility and vision will make possible the construction of more powerful robots. As a consequence, in this temporary scenario the automation of all tasks in all sectors increases, and the high level of integration of robots in the areas of security and defense stands out. On a long run horizon, with the year 2050 as a reference, the presence of robots will be common and the humansmachines relationship will be strengthened in an exceptional way. If the levels of automation projected by experts are reached, this will redefine the meaning of human activity, in a context in which robots may become, in the words of Hans Moravec, everyday companions in leisure or work. In this view, technological innovation will become increasingly important. The activities capable of being rearranged or automated in the services sector will tend to be carried out by robots (as in the industrial sector). And each time it will be more significant the social and political debate will focus on two questions: firstly, how to reconstruct the security, quality and quantity of jobs in more technology-savvy economies (note however that in a global context, the technological innovation is not limited to the most advanced countries: the manufacturing processes in the whole planet tends to incorporate the highest level of automation possible); secondly, how to manage the impacts derived from a higher interaction between humans and machines in all areas of daily life. In this sense, the experience in recent years points out three trends: a higher automation of industrial jobs in the OECD, increased productivity and a higher socializing with robots and automatic systems in daily life and leisure. As the International Labor Office points out, the processes in progress should not be taken as a fixed destiny: they should be seen as a possible opportunity to reopen the social policy [36]. In this sense, the research on the impacts of robotic and advanced automation allows us to examine in depth changing trends. The interaction with household and services robots, the industrial automation, the convergence or symbiosis between people and machines call for developing prospective studies to help decision-making. As we saw above the aspects unearthed by such prospective analyses can go to the heart of our assumptions, preferences, and choices about the social context which generates these technologies and which is in turn shaped by them. References
[1] U. Beck, Power in the Global Age, Polity Press, Cambridge, 2005. [2] M.C. Roco, W.S. Baingridge (Eds.), Converging Technologies for Improving Human Performance. Nanotechnology, Biotechnology, Information Technology snd, Cognitive Science, National Science Foundation (USA), Arligton, Virginia, 2002. [3] H. Moravec, Robot. Mere machine to transcendent Mind, Oxford University Press, New York, NY, 1999. [4] M.L. Silver, Consecuencias para el futuro de las innovaciones en biotecnologa, in: J.F. Tezanos (Ed.), Los impactos sociales de la revolucin cientfico-tecnolgica, IX Foro sobre Tendencias Sociales, Sistema, Madrid, 2007, pp. 530. [5] R. Kurzweil, The Singularity is Near. When Human Transcend Biology, Viking Penguin Group, New York, NY, 2005. [6] B. Joy, Why the future doesn't need us?, Wired Magazine, 8.04 (2000) (http://www.wired.com/wired/archive/8.04/joy.html)
1200
[7] G. Gaskel, A. Allansdottir, N. Allum, C. Corchero, C. Fischler, J. Hampel, J. Jackson, N. Kronberger, N. Mejlgaard, G. Revuelta, C. Schreiner, S. Stares, H. Torgersen, W. Wagner, Europeans and Biotechnology in 2005: Patterns and Trends. Eurobarometer 64.3, European commission's Directorate General for Research, Brussells, 2006. [8] P. Singer, A. Daar, Avoiding frankendrugs, Nat. Biotechnol. 18 (2000) 12251235. [9] S. Jasanoff, Designs on Nature. Science and Democracy in Europe and the United States, Princenton Universitary Press, Princenton and Oxford, 2005. [10] National Institute of Science and Technology Policy (NISTEP), The Seventh Technology Foresight. Future Technology in Japan toward the Year 2030, Science and Technology Foresight Centre, Ministry of Education, Culture, Sports, Science and Technology, Tokyo (Japan), 2001. [11] P. Goldschmidt, Scientific inquiry or political critique? Remarks on Delphi assessment, expert opinion, forecasting and group processs by H. Sackman, Technol. Forecast. Soc. Change 7 (3) (1975) 195213. [12] H.A. Linstone, M. Turoff (Eds.), The Delphi Method. Techniques and Applications, Addison-Wesley, Reading, MA, 1975. [13] M. Hder, S. Hder (Hrsg), Die Delphi-Technik in den Sozialwissenschaften. Methodische Forschungen und innovative Anwendungen, Westdeutscher Verlag, Wiesbaden, 2000. [14] M. Adler, E. Ziglio, Gazing into the Oracle, The Delphi Method and its Application to Social Policy and Public Health, Jessica Kingsley Publishers, London, 1996. [15] Ministre de l'Enseignemetn Suprieur et de la Recherche (MESR), Enqute sur les technologies du futur par la mthode Delphi, MESR, Paris, 1995. [16] D. Loveridge, L. Georghiou, M. Neveda, United Kingdom Technology Foresight Programme, Delphi Survey, University of Manchester, Manchester, 1995. [17] Institut fr Trendanalysen und Krisenforschung (ITK), Delphi Report Austria, vol. 5: Gesellschafts und Kultur- Delphi I: Die Zukunft der sterreichischen Gesellschaft. Experten-Szenarien fr die Jahre 2003-2013-2028, Bundesministerium fr Wissenschaft und Verkehr, Wien, 1998. [18] Fraunhofer-Institut Systemtechnik und Inovationsforschung (FISI), Delphi 98. Studie zur Globalen Entwicklung von Wissenschaft und Technik. Methoden und Dateband, Bundesministerium fr Bildung, Wissenschaft, Forschung und Technologie, Bonn (Deutschland), 1998. [19] A. Lpez Pelez, Towards a new work pattern? Trends of automation and robotics systems in manufacturing and services, Robotics Newsletter, J. Int. Fed. Robot. 40 (2000) 810. [20] A. Lpez Pelaz, Technology, Encyclopedia of Life Support Systems (EOLSS), UNESCO, Paris, 2004. [21] A. Lpez Pelez, M. Krux, Social Impacts of Robotics and Advanced Automation towards the Year 2010, The IPTS Report (edited by The Institute for Prospective Technological Studies European Commission-), 48, 2000, pp. 3440, 2000. [22] A. Lpez Pelez, M. Krux, Future Trends in Health and Safety at Work: New Technologies, Automation and Stress, The IPTS Report (edited by The Institute for Prospective Technological Studies European Commission-), 65, 2002, pp. 2433. [23] A. Lpez Pelez, M. Krux, New Technologies and New Migrations: strategies to enhance social cohesion in tomorrow's Europe, The IPTS Report (edited by The Institute for Prospective Technological Studies European Commission-), 80, 2003, pp. 1117. [24] S. Watanabe (Ed.), Microelectronics, Automation, and Employment in the Automobile Industry, John Wiley and Sons, Sussex, 1987. [25] A. Lpez Pelez, J.A. Daz Martnez, Science, technology and democracy: perspectives about the complex relation between the scientific community, the scientific journalist and public opinion, Soc. Epistemol. 21 (1) (2007) 5568. [26] International Federation of Robotics (IFR), World Robotics, ONU/IFR, Geneve, 2006, p. 4. [27] SOPEMI, International Migration Outlook, OECD, Paris, 2006. [28] A. Lpez Pelez, M. Krux, New Technologies and New Migrations: strategies to enhance social cohesion in tomorrow's Europe, The IPTS Report (edited by The Institute for Prospective Technological Studies European Commission-), 80, 2003, pp. 1314. [29] D. Levy, Love and Sex with Robots: The evolution of Human-Robot Relationships, Harper, London, 2007. [30] Foundation for the Improvement of Living and working conditions, Third European Survey on working conditions, Dublin, 2001. [31] U. Beck, Un nuevo mundo feliz. La precariedad del trabajo en la era de la globalizacin, Paids, Barcelona, 2000. [32] R.A. Brooks, Flesh and machines: How robots will change us, Pantheon Books, Random House, 2002. [33] W. Mitchell, Me++,The Cyborg Self and the Networked City, MIT Press, Cambridge, 2003. [34] Technological Forecasting and Social Change V73, Issue 2, February 2006, pp.95-127
A. Lpez Pelez, D. Kyriakou / Technological Forecasting & Social Change 75 (2008) 11761201 [35] T. Caplow, American Social Trends, Harcourt Brace Jovanovich, San Diego, 1991. [36] Internacional Labour Office (ILO), Seguridad econmica para un mundo mejor, MTAS, Madrid, 2005, p. 251.
1201
Antonio Lpez Pelez, PhD, is Associate Professor in the Faculty of Sociology and Political Sciences, Universidad Nacional de Educacin a Distancia, Madrid, Spain. His most recent books are Nuevas Tecnologas y Sociedad Actual: el impacto de la Robtica, Madrid, Ministerio de Trabajo y Asuntos Sociales, 2003; Trabajo Social Comunitario: afrontando juntos los desafos del siglo XXI, Madrid, Alianza Editorial 2008 (with Toms Fernndez).
Dimitris Kyriakou works at the European Commission's Institute for Prospective Technological Studies, where he has been responsible for running the refereed techno-economic journal The IPTS REPORT, and as chief economist for the management of the institute's economic advisory group led by Nobel-prize-winner Bob Solow. He has co-authored numerous articles and books on techno-economic issues.

Robots, Genes and Bytes: Technology Development and Social Changes Towards The Year 2020

Încărcat de

Informații document

Titlu original

Drepturi de autor

Formate disponibile

Partajați acest document

Partajați sau inserați document

Opțiuni de partajare

Vi se pare util acest document?

Este necorespunzător acest conținut?

Drepturi de autor:

Formate disponibile

Robots, Genes and Bytes: Technology Development and Social Changes Towards The Year 2020

Încărcat de

Drepturi de autor:

Formate disponibile

Available online at www.sciencedirect.

Technological Forecasting & Social Change 75 (2008) 1176 1201

2010 2010 2010 2025

2015 2025 2020 2025

2015 2020 2020 2025

Delphi study 2005 Trends Security in the projection

We thank one of the anonymous referees for this example.

S-ar putea să vă placă și