What is the initial feedback from IT? Where does the resistance lie?
As you might expect, resistance generally comes from engaging any team late in the process or without sufficient information or sponsorship. Generally, we have found IT teams to be hugely supportive of RPA when it is deployed within IT governance and addresses a challenge that IT is not already addressing through other programs of work.
Information security is generally the IT team we spend most of our time with. RPA provides a new model for understanding and constructing appropriate controls. The thought of a robot performing transactions on an unlocked machine accessing sensitive data has obvious risks. Working with the information security team to propose, review and implement controls to manage these risks is essential to a successful deployment.
How should business leaders message the value of RPA to IT?
RPA has numerous benefits to a business, such as improved quality and consistency, reduced transaction times, business continuity and agility – not to mention the obvious cost savings. That said, it is not the only tool in the tool box and IT teams may have different approaches they are already pursuing to solve the same challenge the business is trying to solve with RPA. Understanding the IT roadmap prior to embarking on any implementation is therefore imperative to avoid conflicting agendas.
It is also essential that a business sponsor with sufficient seniority is identified to allocate project budget and prioritize RPA among other initiatives. We’ve found the key to obtaining sponsorship is to perform a ‘Future of Work Assessment,’ or FOWA, across the area of the business. The FOWA evaluates potential solutions, proposes a Target Operating Model (TOM) and compiles a business case to articulate the value of RPA and the cashable and non-cashable benefits it will bring. Once the investment and benefit are quantified, it’s easy to justify RPA and resources in supporting its implementation.
Why is it important for IT to be involved in implementation?
While RPA is often managed by operations teams to provide a virtual workforce, it is still an IT implementation, and therefore has to be deployed and managed within an IT governance framework so that the risks associated with automation can be effectively managed.
We have seen some organizations take a different route in implementing RPA without the involvement of IT. In every single instance, this has caused additional delay and/or risk to the business, and in the vast majority of cases, has resulted in a lag in adoption or the RPA initiative being shut down altogether.
IT teams tend to be the budget holders for the infrastructure that is put in place (new robots), and are responsible for infrastructure and system availability, up time and recovery. IT teams also tend to hold the licensing and roadmaps for the target applications RPA is automating (office, SAP etc.).
Regardless of which function manages the implementation, for RPA to be successful both operations and IT have to be bought into the initiative and actively involved.
When should IT get involved?
In our experience, it is best to involve IT from the outset. This doesn’t mean heavy involvement – just the socializing of the investigation, which may or may not lead to a business case for RPA. The earlier the engagement, the less risk to the project as you may discover that the application you wish to automate is due for replacement in the following year or that there is already another RPA pilot in your organization that you could leverage.
Once the idea has been socialized, and ideally once the sponsorship is in place, the next step is to perform a FOWA assessment. We have found that a strong business case is far more convincing to leadership than a compromised proof of concept that proves the software can work in your business as it does now in countless other organizations. Once the FOWA is complete, IT needs to be involved in validating the security model and providing the governance within which the deployment can take place. Often the IT team will be required to provide access to non-production environments that mirror the live systems for the purpose of developing and testing RPA.
Once the implementation is complete, IT involvement is more important than ever, as they need to manage any upgrade or change to the systems being automated to ensure continuity of automation.
Where has IT seen value? How have their jobs been made easier?
RPA is a great means to address the projects that IT cannot prioritize. We’ve worked with several CIOs and CTOs who have told us that they now look at RPA in their triaging of investment requests. If the opportunity is not significant enough to make it onto the roadmap, then the IT teams look to see if RPA can provide a lower cost pragmatic solution. This is a great dynamic to create as our clients that deploy RPA do not want to compete with the initiatives to replace systems or upgrade their functionality, rather they want to find a more effective solution than dealing with these problems manually as they do today.
With many projects delivering a typical payback period of less than one year, there is often a case for implementing RPA even when there are longer term strategic solutions for addressing the same challenge.
Another way RPA can be used by IT teams is as a means of prototyping automations which can then be transferred into the underlying applications when stabilized.
How do you see RPA impacting IT day to day in the next five years?
I think RPA will become a more valuable tool to IT departments where they can provide operations teams with a means to provide automated solutions to problems that are not addressed through the IT roadmap.
We do not see it as a means to reduce IT spend or channel away the limited funds IT teams usually have available to them. Instead, we see it as a way to extend the amount of opportunities they can support by enabling operations with tools that have been approved by IT and are managed effectively and securely.
There is also the potential to grow hybrid IT / business roles by bringing the functions closer together.
Automation is not a singular discipline.
A new study suggests that business and IT automation is taking over tasks, not jobs.
The implementation of robotic process automation (RPA) is enabling enterprises to execute business processes 5-10 times faster with an average of 37 percent fewer resources, according to a report released this week by Information Services Group (ISG). However, the productivity gains are not necessarily leading to mass layoffs, but rather the redeployment of employees to handle higher-value tasks and a greater volume of work, according to ISG
“Automation is creating a polar shift in how work gets done,” says ISG partner Craig Nelson. “While in the past humans have been supported by technology, we are now seeing a shift to technology being supported by humans to manage and operate business processes. This shift is eliminating much of the mundane cut-paste-and-compare work that humans manage in the cracks between enterprise systems.”
The initial response to automation improvements is typically positive, says Nelson, as the technology takes over some of the dirty work employees are eager to offload. But then the anxiety can set in. The elimination of tasks can lead to the elimination of low-level roles, says Nelson. After all, the initial business case for automation was based on eliminating work and full-time employees. “However, as leaders have gained more experience, it is clear that robots are good at automating specific discrete tasks, not a person’s entire job,” Nelson says. “The extra capacity generated by automating tasks is being focused on executing more work or higher-value work.”
Rethinking RPA’s value
As CIOs and other leaders gain more experience with RPA, they are now looking at the automation technology within the broader digital transformation of the enterprise. “This entails understanding how RPA can support the digital backbone of the enterprise with automation and then moving to understanding the predictive analytics available with automation, which gives the enterprise greater insights into its business, customers and products,” says Nelson.
Automation can also lead to the creation of new roles. “Longer term, the answer for workers is to embrace the polar shift toward skills required for humans to support technology,” says Nelson. New roles might include working in a robotics center of excellence, supporting automation configurations, process redesign and business digitization. IT tasks like writing scripts, monitoring infrastructure and applications, or providing desktop support are ripe for automation but there will be increased work involving business relationship management, configuring and maintaining automation, change control, and monitoring service strategy, as examples.
“Understanding the broader digital transformational journey and thinking about the human interactions that are required when an enterprise begins to engage its customer digitally puts RPA and job disposition considerations in a different light,” says Nelson. “The opportunity for job creation in this space is yet to be fully understood, but it is certain to create new roles and new jobs that we have not yet envisioned.”
Taking the long view
To date, most corporate leaders have focused on the cost reduction that the application of RPA can enable by reducing reliance on labor and outsourcing. Therefore, some leaders have been eager to eliminate processes and roles as soon as possible. But that’s a shortsighted approach, says Nelson. “The longer-term implications regarding talent retention and employee development are not being adequately addressed as the mad scramble for the cost savings tends to take priority over the impact of automation on the culture of the organization and considerations regarding the journey toward becoming a digital enterprise.”
RPA is typically deployed by line-of-business leaders rather than IT who see it as an easy way to eliminate costs while improving speed, accuracy and auditability. And since there’s no need to program these robots, IT often times is only involved in provisioning the infrastructure and making sure the solution is deployed using the right architecture.
Barely six months ago, President Donald Trump rode a wave of anti-establishment sentiment to the White House. One of his central campaign promises was to bring back jobs to American workers, vowing to “Buy American, Hire American.”
The political merits of this approach can be debated, but there’s something that can’t be denied: Some jobs can’t be brought back, especially those replaced by technology and artificial intelligence. Automation stands apart from the debate around immigration or globalization. Robots can now assemble cars, around the clock, without much help from humans. Machines can write stories for news publishers. It’s simply the way it goes — technology does displace workers – but, perhaps more surprising, is that many people don’t object to artificial intelligence (AI) taking our jobs. In some cases, they actually are in favor of it.
Because there’s so much noise and hype around AI. It’s difficult to determine what matters and how it will impact the future of business.
AI’s potential to boost business outweighs the potential downside of job losses, according to a PWC survey of 2,500 business executives and consumers. But this comes with one key caveat: People want AI platforms that replace humans to provide more affordable solutions and products to the wider population. For example, 80% of respondents say it’s of greater importance to have access to more affordable legal advice than to preserve the jobs of lawyers. And 69% would rather have more affordable, convenient and reliable transportation than preserve the jobs of taxi drivers. Respondents felt that if human jobs are replaced, the AI platform replacing them needs to benefit the wider population.
Businesses are already making the necessary decisions and investments to utilize AI to a greater degree (54% said they are making “substantial investments” in AI). Executives cite the potential for AI to elevate employees from minute, tedious projects to allow them to do more important work, use digital assistants to better manage schedules, and detect data trends to better inform strategy. But to make this all work and not have mass unemployment, everyone needs to be prepared to gain the competencies to work effectively with new technologies, as people will need skills for platforms that may have not even existed just a few years ago. The majority of experts expect technology like AI to create more jobs than it displaces by the year 2025, according to Pew, but these jobs may be completely different from roles that exist today.
What’s most surprising is that executives are willing to trust AI for such important decisions as promotions and salaries: 69% thought an AI platform would be as fair or even more fair as a human in making promotion and salary decisions, but 86% of respondents would still want to talk to a human after a review decision was made by AI, suggesting that people want the intelligence of an AI platform, but paired with the empathy of a human. This theory goes beyond the business world: Even in a day and age when it’s difficult to arrange an appointment with a doctor, the survey found that 77% prefer to visit a doctor in person versus taking an assessment at home with a robotic smart kit. These results suggest that the safest human jobs will be those that involve a person-to-person connection that can’t be easily replaced by machine.
People do have concerns about AI. Privacy is a big concern for people, with 87% citing privacy as a “major concern” for AI, while 23% of respondents believe AI will have serious, negative implications. While people see the positive potential of AI, they also want safeguards to ensure it’s not abused.
Overall, people are more excited about the potential for good than they are worried about the negatives. AI provides the potential to make services vastly more accessible, more affordable, more efficient for everyone, and even more personal. And that’s something that all of us, no matter what side of the political aisle we’re on, should be able to get behind.
The pervasive fear that artificial intelligence (AI) will take over human economic livelihood has been felt in places like the manufacturing sector, as large swaths of the industry automate labor formerly done by humans. However, proponents of machine learning say ultimately AI and robotics will improve the way we do virtually everything, and ultimately create new jobs.
Still, nearly 40 percent of U.S. jobs were slated as a “high risk” for automation by the early 2030s in a March 2017 report by PricewaterhouseCoopers (PwC). While the PwC report acknowledges it’s unlikely all those jobs will be automated for “a variety of economic, legal, and regulatory reasons,” PwC also acknowledges that new tech typically means the creation of new jobs for human workers as well, conceding “the net impact of automation on total employment is therefore unclear.”
Many technologists purport that the new job creation will offset some of the pain of displacement; retraining programs and continuing education opportunities are key to bringing in displaced workers into the new high-tech fold.
“Ever since the industrial revolution, we’ve created technology that in theory has displaced workers, and yet growth continues,” Chris Volinsky, assistant VP or inventive science at AT&T Labs, told Business News Daily. “It’s a displacement of work from more menial tasks to those tasks which require more education and more technology, so work gets displaced, but workers are constantly evolving and being retrained.”
However, the pace of technological growth is so fast that many workers might not find this a truly viable option, said Moshe Vardi, professor of computer science at Rice University and fellow at the Institute of Electrical and Electronics Engineers.
“As AI becomes more effective and complex, the zone of ‘automated jobs’ will continue to widen across industries and verticals,” Vardi said. “Workers are racing against the machines, and to stay ahead of the game, they need to be willing to continually refresh and upgrade their skills. The jobs least likely to be automated are those that combine nonroutine technical skills in combination with people skills.”
What’s on the horizon?
The pace of change has workers both excited and anxious about what the future holds. According to a survey conducted by Atlassian, 87 percent of respondents expect AI to change their jobs by 2020, with 76 percent responding that some or half of their job could be performed by an algorithm or robot. And while 64 percent said they trust AI’s ability to properly complete a task, 80 percent are concerned about a subsequent spike in unemployment.
“If harnessed correctly, AI can become our team’s ‘sixth man,’ moving beyond digital assistants and chatbots, and freeing up time and headspace for us to tackle society’s most complex problems,” Atlassian’s report reads.
“AI is, first and foremost, a tool that makes humans more productive, not unlike a hammer or a steam shovel,” said Manuel Ebert, founder of AI and machine intelligence consultancy summer.ai. “If one human can produce more in the same time, that means we need fewer humans to satisfy the same demand. That is where displacement comes from.”
However, Ebert continued, when productivity increases and costs decrease, oftentimes the demand for those goods and services increases and helps drive the creation of new jobs.
“Think the printing press and books or the assembly line and cars,” he said. “So, the interesting question is where can AI create demand for things that were previously (prohibitively) expensive?” [See Related Story: AI Comes to Work: How Artificial Intelligence Will Transform Business]
The PwC report anticipates a rise in average pretax incomes because of mass adoption but acknowledges “these benefits may not be evenly spread across income groups.”
“There is therefore a case for some form of government intervention to ensure that the potential gains from automation are shared more widely across society through policies like increased investment in vocational education and training,” the report reads.
Others have suggested a universal basic income of some kind, which would essentially offer payments to citizens that could cover necessities like groceries or rent and mortgage payments.
How will the job market transition into automation?
While it is generally agreed that some steps need to be taken to ameliorate the pain of transition, AI proponents like Volinsky argue that the benefits of these technologies far outweigh the negatives. For example, he said, AT&T is utilizing drones and machine learning to expedite inspection and maintenance of cell towers. Instead of sending a worker up, the company now flies drones to inspect the antennae.
“(The drone) flies up with HD video and sends footage back to a technician on the ground to inspect,” Volinsky said. “It might take a half hour to do a full detailed inspection of one of those towers, even though the technician is only interested in certain parts of that video.”
That’s where AI comes in: Machine learning can be used to identify potential problem areas and highlight key points of interest the human technician needs to analyze, Volinsky said. By doing so, it can reduce the half hour task to a matter of minutes, removing the technician’s need to scan through useless pieces of video to find the value.
AI is also making headway in customer service, internal decision-making and the way companies track their customer relationships, to name a few examples. Each of these in-roads represent only the beginning of the AI revolution, Volinsky said, and these tools will be essential as they proliferate.
“I like to think of AI as taking the mundane parts out of peoples’ work and helping humans focus on their real expertise, which is identifying problems and focusing on what else needs to be done,” Volinsky said.
As AI more prominently enters the workforce, humans will need to prepare for continued waves of automation by learning new skills and adapting to a changing economy, while harnessing the capabilities of AI to solve problems that were previously out of reach. A symbiotic relationship between man and machine, then, appears far more desirable than a war for prominence.
The world is changing so rapidly around us that it is difficult to put things in perspective; what is significant, which changes are relevant? Machine intelligence and robotics are giving rise to a whole new breed that utilizes machines both physically and mentally. This transition is bound to change cultures as well as patterns related to work and earning, while also requiring us to learn new things at an ever-increasing pace.
Mechanization of cognitive human abilities
The social and technological evolution of mankind only began a while ago. For more than 200,000 years, the evolution of humans as a species was fairly slow and happened in small steps. However, at the end of the 19th century, we embarked on a journey as a species. So far, this journey has taken us to space, to other planets and to this moment, where we are truly wondering where the machines we have invented will take us and whether we will be able to develop as a species and society in a way that allows as many people as possible to benefit from the wellbeing created by machines. At the same time, we are afraid that machines might take away our importance, jobs and livelihood. Machine intelligence and robotics represent a transformation at least equal to the one started by the steam engine in the late 19th century. At that time, we multiplied our physical strength by utilizing machines. We were able to move from one place to another more quickly and produce things at a faster pace with machines. We created the industrial society that produced a Western standard of living – much like the one we are now enjoying in Finland as well.
The next significant thing we are witnessing is the mechanization of cognitive human abilities such as seeing, hearing, speaking and thinking. Robots used to replace human labor in physical activities, but today artificial intelligence and robotics are able to outperform us in intellectual tasks in terms of speed, quality and results. This means that some of the tasks of traditional highly educated professionals, such as doctors, lawyers and accountants, will be taken over by machines. Robotic Process Automation (RPA), is already available to all organizations and can be used to automate routine processes in areas such as payroll administration. Cognitive computing and artificial intelligence, which is represented by actors such as IBM Watson, enables teaching computers vast amounts of data. Computers learn and are also able to have a natural conversation in our language, but they can process and analyze information many times faster than we can. Even a very simple software robot can be used to automate up to 80% of rule-based manual processes, and a robot works tirelessly 24/7.
Digital and cognitive solutions generate efficiency for organizations
We are on the verge of an era that will force us to abandon old-fashioned conceptions of learning and teaching. Memorizing things is no longer sensible, as everyone can access answers just by asking a computer. When practically anyone can use artificial intelligence to make decisions about matters that used to require decades of practice and studying, we need to change our way of learning and studying new things. For knowledge workers, this is a great opportunity to take a quantum leap in efficiency and change the nature of work so that it becomes more creative and non-mechanistic. Cognitivity combined with new digital learning opportunities can play a decisive role when the basic structures of our society change as digitalization, robotization and machine intelligence become commonplace.
For companies, digital and cognitive solutions enable improved customer service through all channels in an individual and efficient way. With the help of machine intelligence and analytics, self-service can be supplemented with digital assistants capable of, for example, helping select the right products, suggesting better options for your specific needs or offering affordable complementary alternatives. Similarly, people’s learning paths can be supported by utilizing a large amount of data on different learning styles, paces and results. Individual study paths created by a cognitive process and even fully customized study materials (level of difficulty, number of exercises, additional reading, etc.) enable the most efficient learning results and can help organizations take major quantum leaps in efficiency.
Many people are unsure about exactly what machine learning is. But the reality is that it is already part of everyday life.
A form of artificial intelligence, it allows computers to learn from examples rather than having to follow step-by-step instructions.
The Royal Society believes it will have an increasing impact on people’s lives and is calling for more research, to ensure the UK makes the most of opportunities.
Machine learning is already powering systems from the seemingly mundane to the life-changing. Here are just a few examples.
1. On your phone
Using spoken commands to ask your phone to carry out a search, or make a call, relies on technology supported by machine learning.
Virtual personal assistants – the likes of Siri, Alexa, Cortana and Google Assistant – are able to follow instructions because of voice recognition.
They process natural human speech, match it to the desired command and respond in an increasingly natural way.
The assistants learn over a number of conversations and in many different ways.
They might ask for specific information – for example how to pronounce your name, or whose voice is whose in a household.
Data from large numbers of conversations by all users is also sampled, to help them recognise words with different pronunciations or how to create natural discussion.
2. In your shopping basket
Many of us are familiar with shopping recommendations – think of the supermarket that reminds you to add cheese to your online shop, or the way Amazon suggests books it thinks you might like.
Machine learning is the technology that helps deliver these suggestions, via so-called recommender systems.
By analysing data about what customers have bought before, and any preferences they have expressed, recommender systems can pick up on patterns in purchasing history. They use this to make predictions about the products you might like.
3. On your TV
Similar systems are used to recommend films or TV shows on streaming services like Netflix.
Recommender systems use machine learning to analyse viewing habits and pick out patterns in who watches – and enjoys – which shows.
By understanding which users like which films – and what shows you have watched or awarded high ratings – recommender systems can identify your tastes.
They are also used to suggest music on streaming services, like Spotify, and articles to read on Facebook.
4. In your email
Machine learning can also be used to distinguish between different categories of objects or items.
This makes it useful when sorting out the emails you want to see from those you don’t.
Spam detection systems use a sample of emails to work out what is junk – learning to detect the presence of specific words, the names of certain senders, or other characteristics.
Once deployed, the system uses this learning to direct emails to the right folder. It continues to learn as users flag emails, or move them between folders.
5. On your social media
Ever wondered how Facebook knows who is in your photos and can automatically label your pictures?
The image recognition systems that Facebook – and other social media – uses to automatically tag photos is based on machine learning.
When users upload images and tag their friends and family, these image recognition systems can spot pictures that are repeated and assigns these to categories – or people.
6. At your bank
By analysing large amounts of data and looking for patterns, activity which might not otherwise be visible to human analysts can be identified.
One common application of this ability is in the fight against debit and credit card fraud.
Machine learning systems can be trained to recognise typical spending patterns and which characteristics of a transaction – location, amount, or timing – make it more or less likely to be fraudulent.
When a transaction seems out of the ordinary, an alarm can be raised – and a message sent to the user.
7. In hospitals
Doctors are just starting to consider machine learning to make better diagnoses, for example to spot cancer and eye disease.
Learning from images that have been labelled by doctors, computers can analyse new pictures of a patient’s retina, a skin spot, or an image of cells taken under a microscope.
In doing so, they look for visual clues that indicate the presence of medical conditions.
This type of image recognition system is increasingly important in healthcare diagnostics.
8. In science
Machine learning is also powering scientists’ ability to make new discoveries.
In particle physics it has allowed them to find patterns in immense data sets generated from the Large Hadron Collider at Cern.
It was instrumental in the discovery of the Higgs Boson, for example, and is now being used to search for “new physics” that no-one has yet imagined.
Similar ideas are being used to search for new medicines, for example by looking for new small molecules and antibodies to fight diseases.
The focus will be on making systems that perform specific tasks well which could therefore be thought of as helpers.
In schools they could track student performance and develop personal learning plans.
They could help us reduce energy usage by making better use of resources and improve care for the elderly by finding more time for meaningful human contact.
In the area of transport, machine learning will power autonomous vehicles.
Many industries could turn to algorithms to increase productivity. Financial services could become increasingly automated and law firms may use machine learning to carry out basic research.
Routine tasks will be done faster, challenging business models that rely on charging hourly rates.
Over the next 10 years machine learning technologies will increasingly be part of our lives, transforming the way we work and live.
As part of techUK’s AI Week, Alastair Bathgate, Group CEO at Blue Prism has provided a blog on ‘Friend not Foe: Robotic workforces could boost – not break – our working world’
Pressure is mounting on us, wherever we work to outperform peers, fend off threats from new challengers and remain competitive. That’s modern life – but each of us is only human. We can’t do everything at once. Some things we just don’t want to do, maybe even ought not to do. In any event, every now and then something’s got to give.
In the past 20 years, Business Process Outsourcing cast itself as the saviour of productivity – sometimes even of corporate sanity. Companies large and small took a long hard look at themselves, at the places where they added value (and where they leaked it) and looked to outside talent to work harder, smarter, faster.
Today, that outside talent alone isn’t enough. The future of work is changing at a faster pace than any of us have ever known. Virtually every company on earth is now looking at ‘what’s next’ – and increasingly that’s software robots.
Robots attract a great deal of emotion. We imagine sci-fi figures from TV and film. We fear replacement, not emancipation, from machines much better at work than we are. But, just like the fears that accompanied BPO in its infancy, the truth and potential around software robotics is far more exciting than this imagery allows.
Robotic Process Automation (RPA) is the game-changer. We’ve all seen industrial robots make waves in manufacturing, deliver efficiency in supply chains and improve product quality. We are now seeing software robots do the same elsewhere.
Software robots put process where process belongs, inside smart systems. They free each of us up to apply our human brains to tasks far more valuable. Let’s face it: traditional IT systems can often creak under the simplest of pressures. There are gaps, flaws, missing links – and all too often human intervention isn’t the best way to mitigate for those failings.
Smart software robots are far better suited to the task. Rather than have employees stretch time and resource to marginally improve an already broken system, what if you could simply add extra robotic capacity to do all that for you? Robots that feed on process, scalability and compliance.
But that’s just the beginning. Imagine a robotic ecosystem where users can leverage best-of-breed solutions for AI, cognitive and cloud technology. Consider the benefits to an a la carte menu of services and capabilities that would let you free up the creativity of people to do more valuable tasks: seek out that new market, revitalize sales, spot new opportunity and double output with virtually zero cost.
Doesn’t this sound like good news for business, not bad? Couldn’t it be good news for productivity, for people and for process? We think so. Increasingly the biggest players in business think so too. Perhaps it’s time we put sci-fi to one side and gave robotics a fresh look…?
Every few months it seems another study warns that a big slice of the workforce is about to lose their jobs because of artificial intelligence. Four years ago, an Oxford University study predicted 47% of jobs could be automated by 2033. Even the near-term outlook has been quite negative: A 2016 report by the Organization for Economic Cooperation and Development (OECD) said 9% of jobs in the 21 countries that make up its membership could be automated. And in January 2017, McKinsey’s research arm estimated AI-driven job losses at 5%. My own firm released a survey recently of 835 large companies (with an average revenue of $20 billion) that predicts a net job loss of between 4% and 7% in key business functions by the year 2020 due to AI.
Yet our research also found that, in the shorter term, these fears may be overblown. The companies we surveyed – in 13 manufacturing and service industries in North America, Europe, Asia-Pacific, and Latin America – are using AI much more frequently in computer-to-computer activities and much less often to automate human activities. “Machine-to-machine” transactions are the low-hanging fruit of AI, not people-displacement.
For example, our survey, which asked managers of 13 functions, from sales and marketing to procurement and finance, to indicate whether their departments were using AI in 63 core areas, found AI was used most frequently in detecting and fending off computer security intrusions in the IT department. This task was mentioned by 44% of our respondents. Yet even in this case, we doubt AI is automating the jobs of IT security people out of existence. In fact, we find it’s helping such often severely overloaded IT professionals deal with geometrically increasing hacking attempts. AI is making IT security professionals more valuable to their employers, not less.
In fact, although we saw examples of companies using AI in computer-to-computer transactions such as in recommendation engines that suggest what a customer should buy next or when conducting online securities trading and media buying, we saw that IT was one of the largest adopters of AI. And it wasn’t just to detect a hacker’s moves in the data center. IT was using AI to resolve employees’ tech support problems, automate the work of putting new systems or enhancements into production, and make sure employees used technology from approved vendors. Between 34% and 44% of global companies surveyed are using AI in in their IT departments in these four ways, monitoring huge volumes of machine-to-machine activities.
In stark contrast, very few of the companies we surveyed were using AI to eliminate jobs altogether. For example, only 2% are using artificial intelligence to monitor internal legal compliance, and only 3% to detect procurement fraud (e.g., bribes and kickbacks).
What about the automation of the production line? Whether assembling automobiles or insurance policies, only 7% of manufacturing and service companies are using AI to automate production activities. Similarly, only 8% are using AI to allocate budgets across the company. Just 6% are using AI in pricing.
Where to Find the Low-Hanging Fruit
So where should your company look to find such low-hanging fruit – applications of AI that won’t kill jobs yet could bestow big benefits? From our survey and best-practice research on companies that have already generated significant returns on their AI investments, we identified three patterns that separate the best from the rest when it comes to AI. All three are about using AI first to improve computer-to-computer (or machine-to-machine) activities before using it to eliminate jobs:
Put AI to work on activities that have an immediate impact on revenue and cost. When Joseph Sirosh joined Amazon.com in 2004, he began seeing the value of AI to reduce fraud, bad debt, and the number of customers who didn’t get their goods and suppliers who didn’t get their money. By the time he left Amazon in 2013, his group had grown from 35 to more than 1,000 people who used machine learning to make Amazon more operationally efficient and effective. Over the same time period, the company saw a 10-fold increase in revenue.
After joining Microsoft Corporation in 2013 as corporate vice president of the Data Group, Sirosh led the charge in using AI in the company’s database, big data, and machine learning offerings. AI wasn’t new at Microsoft. For example, the company had brought in a data scientist in 2008 to develop machine learning tools that would improve its search engine, Bing, in a market dominated by Google. Since then, AI has helped Bing more than double its share of the search engine market (to 20%); as of 2015, Bing generated more than a $1 billion in revenue every quarter. (That was the year Bing became a profitable business for Microsoft.) Microsoft’s use of AI now extends far beyond that, including to its Azure cloud computing service, which puts the company’s AI tools in the hands of Azure customers. (Disclosure: Microsoft is a TCS client.)
Start in the back office, not the front office. You might think companies will get the greatest returns on AI in business functions that touch customers every day (like marketing, sales, and service) or by embedding it in the products they sell to customers (e.g., the self-driving car, the self-cleaning barbeque grill, the self-replenishing refrigerator, etc.). Our research says otherwise. We asked survey participants to estimate their returns on AI in revenue and cost improvements, and then we compared the survey answers of the companies with the greatest improvements (call them “AI leaders”) to the answers of companies with the smallest improvements (“AI followers”). Some 51% of our AI leaders predicted that by 2020 AI will have its biggest internal impact on their back-office functions of IT and finance/accounting; only 34% of AI followers said the same thing. Conversely, 43% of AI followers said AI’s impact would be greatest in the front-office areas of marketing, sales, and services, yet only 26% of the AI leaders felt it would be there. We believe the leaders have the right idea: Focus your AI initiatives in the back-office, particularly where there are lots of computer-to-computer interactions in IT and finance/accounting.
Computers today are far better at managing other computers and, in general, inanimate objects or digital information than they are at managing human interactions. When companies use AI in this sphere, they don’t have to eliminate jobs. Yet the job-destroying applications of AI are what command the headlines: driverless cars and trucks, robotic restaurant order-takers and food preparers, and more.
Make no mistake: Automation and artificial intelligence will eliminate some jobs. Chatbots for customer service have proliferated; robots on the factory floor are real. But we believe companies would be wise to use AI first where their computers already interact. There’s plenty of low-hanging fruit there to keep them busy for years.