One of the overriding concerns about technology today is that artificial intelligence and robots will soon take many of our jobs away. Even technology leaders are stoking those fears.
Only a few weeks ago, Tesla Motors Inc. Chief Executive Elon Musk talked about the end of the workforce as we know it, alluding that a Marxian future was on its way replete with added leisure time for us while bots take care of the hard work. Musk said that governments should get ready for this and should start thinking about concepts such as a Universal Basic Income.
In a decidedly less utopian development, 60,000 human jobs were replaced last year with robots at just one company, Taiwanese electronics manufacturer Foxconn Technology Group. And Ed Rensi, a former McDonald’s chief executive, said after experiencing the power of bots at an expo in May that if service staff kept yammering on about higher wages, the fast food industry would soon see “a job loss like you can’t believe.”
But all those developments mask a more likely scenario in which humans and bots work side by side for a long time to come. That’s according to Mihir Shukla, cofounder and CEO of software bot producer Automation Anywhere. The company employs robotic process automation, the use of software applications that replicate how people interact with a computer system, to automate mundane services such as data entry, document verification, invoice processing, payroll administration and complaint management.
Shukla anticipates employing a “digital workforce” of 3 million such bots, already used by companies such as AT&T and ANZ Banking Group, by 2020. That, he declares “in a manner of speaking” could make the company the world’s largest employer, somewhat like Uber Technologies Inc. could be seen as the world’s largest taxi company even without owning taxis or employing drivers.
Although Shukla concedes that bots will replace many jobs, he thinks that even more, they will “work closely with the human workforce and augment talent instead of replacing it.” Shukla recently answered a few questions to help us better understand the future of automation and the implications for the workforce. This is an edited version of the email exchange:
Q: What are some examples of how bots such as Automation Anywhere’s are replacing human jobs?
A: A bot workforce can replace many mundane, repetitive tasks which can be automated, freeing people to create, think, discover, and ultimately build great companies. There are several examples of this today:
- A major global bank automated its banking transactions for higher efficiency and productivity and automated reporting.
- A major retailer used robotic process automation to significantly enhance factory to store to consumer turnaround time.
- A major transportation and logistics provider used automation to audit, track and have real-time analytics of every bill that is generated by the system.
- A local U.S. county health and human services agency in California used automation and bots for a complex document verification process which involved document processing and digitization on its legacy system. Results were significant improvements in turnaround time, streamlined rudimentary processes, document error reduction, reduced risk and overall cost savings.
- One final example is the Australian bank ANZ. Its goal was to be more agile in managing workload variability, reducing risk from error and helping employees acquire new skills. In the first six months, ANZ deployed about 100 “digital workers” (bots) and the bank is adding 100 more each quarter. Cost savings are over 40 percent and response by employees to opportunities to learn new skills and improve their productivity has been overwhelmingly positive.
Q: Which jobs you foresee being replaced first and which jobs will be in line for elimination in coming years?
A: If you look at a funnel of talent, jobs that could be impacted are routine rule-based ones, followed by routine cognitive jobs, followed by heavily data-driven cognitive work:
- Back-office work is a good example of routine rule-based jobs.
- Back- and mid-office and a lot of front office work is a good example of many routine cognitive jobs, like answering calls or processing mortgages.
- Examples of heavily data-driven cognitive are certain types of legal work. For example, if you have to figure out if a 200-page legal document meets all the requirement of certain government regulation described in 10,000 pages of documents, it requires data-driven cognitive work.
Q: Tesla Motors CEO Elon Musk, Baidu Chief Scientist Andrew Ng and President Barack Obama have each mentioned the idea of a basic income as more people could lose jobs. How do you see human collaboration with technology as opposed to mass unemployment as a result of automation?
A: I believe there will be a balance. Yes, automation can automate any process out there. Yes, cognitive technologies can mimic human action through learning. Automating a particular task, so that it can be done more quickly or cheaply, increases the demand for human workers to do the other tasks around it that have not been automated.
Imagine a construction site today. One could argue that our use of machines in construction have reduced jobs but without construction automation, there would be no Manhattan. There are in fact more higher-paying jobs in construction and other industries, because of automation. Use of automation in construction has led to much higher demand of raw materials creating more jobs and enabling today’s Fortune 100 companies to scale rapidly, in turn creating higher-paying jobs and today’s sophisticated infrastructure.
The same will be true yet again. The modern workforce will consist of humans and bots working side by side, each doing what they do best. And not too different from construction site of today, it will become the norm leading us to new heights of achievement, job satisfaction and prosperity.
Q: What are your thoughts on the possibility of a Universal Basic Income?
A: When considering the possibility of basic income, there are three pillars to explore in order to make this transition smooth:
1) Leaders of private companies must engage now. Current workforces have a wealth of knowledge regarding their respective business domains along with an understanding of working with fellow co-workers; this when combined with digital skills, will bring the new age of hyper-productive workforce. Leaders must begin training their workforce with digital and automation skills. It is possible. AT&T is showing companies how it is done.
2) Educational institutes must change how we teach and what we teach. Some of our Victorian-era education systems and ways of thinking need reinventing. Skills that are long overdue for automation should not be taught anymore. The preparation of a digital workforce begins in the classroom with solution focused, design thinking, automation approaches, etc.
3) Transformation at a global scale. Recognition by governments across the globe that this will happen, if not in the U.S., then in other countries, is key to a smooth transition. Policies that incentivize employee training, adding higher value skills and jobs for future growth, will go a long way in helping shape the digital workforce of the future.
Q: President-elect Donald Trump has discussed bringing jobs back to the U.S. Meanwhile, we are being told machines will be replacing many more of these jobs. Do you think politics is somewhat behind tech? Is the government prepared for great changes relating to advanced technologies? Could the public become fearful of such technology, a kind of Luddite reaction to the technological revolution?
A: There is a symbiotic relationship between government and the private sector. The solution is a three-way solution with government playing a very important role. The transformation has just begun. Organizations are creating and deploying a digital workforce today. Some are celebrating with cake every time a new process is automated.
As it relates to resistance, no technology advancement that causes a paradigm shift has been made without resistance. I don’t see a Luddite crisis and the reason why is that the big difference between then and now is the speed of information and adoption. Think about it, we did not have iPhones 10 years ago, but today a billion of us or so walk around with these phenomenal computing devices in our pocket. We did not see a Luddite crisis when 411 was replaced by Siri or Yellow Pages.
Crises happen when unexpected events are thrown upon us. In this case, we know this transformation is coming. There is time to plan and manage the process successfully, provided we start now. There has never been a better time to impact human lives, improve standard of living and solve many problems that have plagued human existence for decades. This should be seen as an opportunity.
Q: The leading tech companies and tech researchers this year decided to form an AI ethics committee, Partnership on AI. Do you have an opinion on what ethical concerns could arise?
A: The history of humanity has shown that at times when enormous power is available, ethics become more important than ever. With enormous power of AI technologies on our radar, ethics are of high importance and we need to decide how to harness this power to improve lives of all people, and make the world a better place.
Ethics plays a huge role in the transformation to a world of tomorrow. It is likely that a few companies that control machines could wield inordinate amount of control. How will they transform the world? How will they balance their ethical obligations vs their self-interests? What will the legal framework look like? Terms like “too big to fail” come to mind.
These are questions we have to wrestle with as a global community and in fact, the definition of a community could change as well. Case in point are the changes we are seeing today across the globe against or for globalization, such as the Brexit vote and the U.S. election of populist Donald Trump.
Q: Do you think technology companies must share some of the responsibility if many people are made unemployed or unemployable? How might this responsibility manifest as action?
A: Thinking about who pays for it before a problem even arises, or is understood, only prevents us from dealing with it. Taxing oil companies for pollution may be part of the solution, but that isn’t a long-term solution to the problem we faced. We had to invent new technologies, uncover alternate sources, and create government policies to encourage innovation on energy.
Similarly, we must use human ingenuity to create new jobs, new solutions, and new ways that work can be done and how work is defined.