What jobs will be impact by Robotics and AI ?
- Mark Skilton
- Sep 4, 2015
- 5 min read
My current research into the number of robot “jobs” there already are in the UK economy and how many jobs have been replaced by robots needs to consider both physical robotics such as material handlers and self drive cars replacing “Human bodies” and the more pervasive Artificial Intelligence algorithms that are in automated tasks replacing “Human heads”. The current corporate financial reporting of human headcount does not take into account activities removed or created by automation but can manifest in the percentage restructuring changes reports in some industries. Notably in cloud computing and in manufacturing and in banking, three sectors that have been heavily disrupted by technology and the shift to automated services. Headcount reductions of human workforces in these cases have seen 10 to 30% reductions in these industries particularly in front desk services and in administrative functions as well as sales and operational staff. These have been absorbed back into the economy as they take up new posts and job roles in the same industry or related , but importantly, broadly the same skills. The issue is when these become replaced and eroded by automation beyond where human reskilling does not create new Human jobs. Evidence of research so far suggest there is a net loss of human jobs in the longer term and a potential restructure of the economy.
INFOGRAPHS
Analyzing the shift in Human work to robotic AI It is not hard to visualize how a typical human day that is spending 1 hour travelling, 8 hours working, 1-2 hours eating and household activities, 2-3 hours social activity could involve 20 to 30% routine activity that could be automated from selecting , ordering and searching with the assistance of computation. But it is the professional jobs as well as the lower skilled tasks that could be under threat.
The issue of the “diminishing need for human labor” from Artificial Intelligence, Employment and Income has been an old issue debated back even in 1984 AI Magazine article by Nils Nilsson (1) and further back in 1933 by the famous economist, John Maynard Keynes, who suggested that technology advancement would ..”in the long run solve mankind economic problem” (2). These argue the potential of technology to rapidly increase productivity and therebycreate more jobs while others said it would remove more jobs due to the advanced robotics and AI removed the need for human intervention. In summary, this is the creation of new skills and human jobs augmented by technology; to reduction of physical human tasks and jobs, and the reduction of mental human tasks and jobs. The overall arching question is if automation then created a “new economy” that is for the betterment of all or reduced the quality of life for those impacted by these forces. Hence it is a societal impact and as much a political and economic one as it is technological.
Just identifying human jobs that can be replaced from an analysis of the social and material dexterity has tended to focus on repetitive tasks or service exchange tasks that don’t require a high degree of variation. These include administration, consumer sales, and service economy functions that have been significantly impact already.
However this is only part of the picture as advances in computing algorithms and robotics are moving into semi-skilled and high skills areas of professional that require judgement and higher functional reasoning. Jobs that are combining both physical dexterity and mental creativity remain further away from the computerization curve today such as engineering, education and management organizational as these involve both actuator movement , the realm of robotics, and advanced contextual thinking that takes analytics and creative thinking together in sense making. There are increasingly several examples of complex algorithm in Bio research for example that can be augmented with computing such as Cancer research, brain functional simulation. likewise in engineering design to forecasting, the use of High performance computing (HPC) is pushing areas of cloud and networked clusters to empower new hybrid research models for delivery. These are not necessarily replacing humans , rather creating new capabilities to collect, aggregate, synthesize new models. That said medical robotics is another field there surgeons may be augmented with Augmented Reality AR and Virtual reality VR that are beginning to blur the boundaries a little. It’s all developing very fast and partly the reason behind the UK Government Whitehall paper recently to explore the impact on human jobs from robotics and artificial intelligence .
The issue is the cost / performance price point of these technologies to replace humans but also the advances in capability beyond human capacity to do things that humans cannot do with a reasonable time space. I’ve adapted a research model to illustrate these predictions
Overall the impact on the number of jobs could be significant over the next ten to twenty years.
Unmanned systems – for example Drones to Warehouse automated material movement to smart stock control systems. These could replace 50-70% of material logistics human tasks
Smart sensor-actuator systems – for example smart environment sensor networks, robotic drivers, connected self-driving vehicles These could replace 20-30% Transport human tasks
Biological Emulation and enhancement– for example neural networks and bioinformatics for assisted medical diagnostics for example and translation medical research. These could replace 10-20% of professional diagnostic tasks Further advances in creative intelligence and leading to a further advances in 10-20% field service human tasks augmented by these capabilities
Contextual aware computing – examples like avatars and mobile assistants to online multi-sided marketplaces that combine different products and services to optimize experiences. This could be a potential massive change of 30-40% of human tasks that today manage these “interfaces” from different vertical industries services into automated horizontal services.
Computational control – examples such as massive image and data processing systems to simulate and create new big data insights impossible by human conventions. This could create new potential but also replace 10-30% of marketing and intelligence gathering human tasks
Augmented and Virtual Reality – examples include 360 video and interactive VR such wearables, to video interactive gaming that simulate “real life” to aid and help locations and services. This could also create new capabilities but alsoimpact 20-30% jobs that hitherto handling physical enquiries and interaction such as service functions.
Who outsources the outsourcers ?
These are just the initial stages, the impact on industry vertical markets could be profound and the ability to engineer a new economic model and work force of the future is at stake. Many tasks that may be out tasked to others or automated could themselves be outsources to smart algorithms and machines.
It is also worth putting it in perspective that the rate of digital data is exploding with forecasts of “billions of connected objects”. The use of mobile services and social networks likewise enabling social network insight into TV viewing habits to advance sin scenario planning for better predictions to speed up and better cost effect medical, financial services to municipal services. But we have only scratched the surface of “digitizing “ everything we touch and use. Remarkably we still have many activities that within a day may be tracking by sensors in our smart phone or street surveillance but have yet to be joined up. Many everyday objects are still not “smart” with connected technology beyond the TV and building services; but these are rapidly changing in the coming storm of the AI wave.
I think it is also important to balance with so call threats with the other sides of the potential for enhanced human life and to offer higher quality service to a larger population that may be disadvantaged or unable to benefit from these advances. It’s a moral and social question of responsibility as well as an economic and technical ethics control issue.
Artificial Intelligence, Employment and Income, Nils Nilsson, Artificial Intelligence Center SRI International, Menlo Park, California 94025 1984 The AI Magazine. http://ai.stanford.edu/~nilsson/OnlinePubs-Nils/General%20Essays/AIMag05-02-002.pdf
Keynes, J. M. (1933) Economic Possibilities for Our Grandchildren (1930). Essays zn Persuasion. London: Macmillan and co.
The future of employment how susceptible are jobs to computerization? , Carl Frey, Michael Osborne. The Oxford Martin Institute, UK, September 2013 http://www.oxfordmartin.ox.ac.uk/downloads/academic/The_Future_of_Employment.pdf
Comments