Shyam's Slide Share Presentations


This article/post is from a third party website. The views expressed are that of the author. We at Capacity Building & Development may not necessarily subscribe to it completely. The relevance & applicability of the content is limited to certain geographic zones.It is not universal.


Sunday, October 1, 2017

Reshaping Business With Artificial Intelligence 3 10-01

7. What to Do Next

Is AI just an element of a company’s overall digital transformation — or does AI require new approaches? On the one hand, AI presents many of the same issues and challenges as other digital technologies, and companies can build in many ways on their digital and analytics programs. However, AI also has distinctive features.

Ensure customer trust. AI capabilities are similar to many digital initiatives that depend on both customer data and customers’ trust that the company will respect and safeguard their personal data. Ensuring that AI is trustworthy is different from other data-dependent digital initiatives, however, in several ways. First, managers may not be able to explain exactly how a customer’s personal data is being used to produce a certain outcome from an AI product. The inner workings of some machine-learning programs are opaque. Second, a growing number of AI systems are able to mimic human agents, putting the onus on managers to clearly communicate to customers whether they are engaging with machines or human agents in a given setting. Third, some AI systems are able to assess emotions and discern quite personal details — at a distance. This capability creates new information management issues, including which employees have access to such information and under what circumstances.

Perform an AI health check. This has some similarities with digital health checks, from applications across processes to enabling infrastructure, technical skills, agile processes, and a fail-fast atmosphere. As with many digital initiatives, success with AI depends on access to data sources, be they existing internal or external data or investments in data infrastructure. Big companies may well have the data they need, but if it is fragmented and siloed, this significantly constrains strategy development and progress. Unlike other digital initiatives, an AI health check involves an assessment of the skills necessary to properly execute the training of AI, from first nurturing the system to become intelligent all the way to continuing to learn after deployment. This is both new and decisive — and a capability most companies need to build themselves. Off-the-shelf AI programs are likely to be limited in their capability and effect.

Brace for uncertainty. The adage “No idea is born good; you have to nurture it over time” applies to AI as well as to digital technologies — only more so. Companies often prioritize their initiatives by estimating the value of, and time required for, establishing a process or offering. But hard estimates are particularly difficult with AI. As a consequence, experimentation and learning with AI can take much longer than other digital initiatives, with a higher variability of success and failure. Managers need to brace themselves for more uncertainty, and this affects how effective they are at prioritizing projects and investments.

Adopt scenario-based planning. Like digital, AI has the potential to shift the ways in which businesses generate value — in multiple markets, processes, and functions. AI requires even more radical thinking, as it affects knowledge- and judgment-based professions, and the new entrants in markets could be machines. Thus, companies need to think even more expansively about their businesses, build cohesive future scenarios, and test the resilience of their directional plans against such scenarios.6 This kind of scenario-based planning can also sharpen the ability to recognize events that could trigger large effects on their business.

Add a workforce focus. AI stands to create significant unease, since even the most knowledgeable expert has difficulties specifying how programs will play out, which functions or processes should be off limits, or where AI should stop. The threat to jobs and careers in their current form is real and can easily lead to employee anxiety and unrest. Establishing a clear focus and work plan for AI initiatives — where they will be pursued and how, including regular communication, education, and training — should be a component of an AI program. Attracting and developing people who combine both business and technical skills will be critical, as will the ability to deploy cross-functional teams, requiring both individual and organizational flexibility.

8. The Way Forward: Implications for the Future

The adoption of AI may have profound effects on the workplace, value creation, and competitive advantage. Beyond the near term, how should companies prepare for these changes?

The Future of Work

As AI is increasingly applied to knowledge work, a significant shift will likely take place in the workplace, affecting many jobs in the Western middle class. Contrary to recent dire predictions about AI’s effect on employment, our survey suggests cautious optimism. Most respondents, for example, do not expect that AI will lead to a reduction of jobs at their organization within the next five years. Nearly 70% also said they are not fearful that AI will automate their own jobs. By a similar margin, respondents hope that AI will take over some of their presumably boring and unpleasant current tasks. However, respondents overwhelmingly agree that AI will both require employees to learn new skills within the next five years and augment their existing skills. (See Figure 12.) Taken together, these portend adjustment, not annihilation. “Even with rapid advances,” says Erik Brynjolfsson, Schussel Family Professor at the MIT Sloan School of Management, “AI won’t be able to replace most jobs anytime soon. But in almost every industry, people using AI are starting to replace people who don’t use AI, and that trend will only accelerate.

Figure 12
Organizations suggest cautious optimism about AI’s effect on the workforce in the next five years.

Shifting Value Creation

Where will AI create, destroy, or shift economic value?

Consider the health care industry, one of the largest and most resilient sources of economic activity in the world. Health care spending makes up one-sixth of the U.S. economy, and on average, about one-tenth of the economies of Organisation for Economic Co-operation and Development (OECD) member nations. AI is already altering the health care value chain: Machines read diagnostic images, surgeons rely on robots, and an ever-increasing number of real-time medical devices contribute and communicate data to improve preventive and chronic care.

While AI may create value within an industry, it is far from clear exactly which organizations will see their fortunes rise and which will see decline. When IT vendors, medtech companies, radiologist networks, hospitals, specialized startups, and even insurance companies all strive to take advantage of AI to improve and lower the costs of diagnostics, the effects of AI will likely be uneven.

It’s too early to tell which types of organizations may benefit from AI in health care. But if regulatory concerns can be worked out, the industry has numerous sources of detailed data. And as Marcus Winter, head of reinsurance development at Munich Re Group, remarks, “In today’s world, with the proliferation of Big Data, there are precious few exclusive data sets. Most of the time, we can triangulate what we need to know via other sources.” In other words, the combination of data and AI algorithms create the possibility of new and more effective workarounds. For example, when diagnostic imaging is unavailable, an evermore accurately analyzed sample of blood or other body fluids might help with diagnosis. As a result, shifts in value creation are difficult to predict.

Building Competitive Advantage

Managers expect significant improvement in performance of current processes or products from AI. Many companies are focused on addressing those. However, mere improvement does not create a sustainable competitive advantage — when everyone finds the same efficiencies, only the baseline shifts. For AI to become a prominent feature in future strategies, companies must figure out how humans and computers can build off each other’s strengths to create competitive advantage. This is not easy: Companies need privileged access to data — which, as we’ve seen, many do not now have. They must learn how to make people and machines work effectively together — a capability relatively few Pioneers have at present. And they need to put in place flexible organizational structures, which means cultural changes for both company and employee.

Just about any company today needs a plan with respect to AI. Most do not have one, and those that have been slower to move have some catching up to do. Those that continue to fall behind may find the playing field tilted evermore steeply against them.

9. Appendix: Work in the Longer Term

Survey respondents and most interviewees both expect big changes from AI in the next five years. But the more dramatic effects of AI may occur within 10 to 20 years. What can we expect in that time frame?

Automation of Tasks ≠ Automation of Jobs. History shows that jobs evolve as tasks shift. BP’s Ahmed Hashmi says the company’s engineers used to spend a lot of time hunting for data to put together their reports, but “now that’s all automated. We’ve got a data lake, which gives engineers ready access to all the data. We employ the same number of engineers, but they’re improving the business rather than searching for data to get ready to improve the business.” In other words, extrapolation from the automation of repetitive tasks to the automation of jobs in a high tech industry is risky business.

AI as Job Creator. Increased organizational reliance on AI will create new needs as it meets current needs. The job of an insurance underwriter, for example, tops many “most endangered species” lists. However, AI simultaneously expands the universe of insurable events. And, as James Platt, chief operating officer of Aon Risk Solutions, has said, “Many things that people would like to insure themselves against, such as brand and reputational risks or wider cybersecurity coverage, are ‘uninsurable’ today. There is simply no one offering an insurance option.” As new methods of assessing risks become available, underwriters can start offering such new services. Missy Cummings, director of the Humans and Autonomy Laboratory at Duke University, puts it this way: “What we often don’t think of are the jobs that are created as other new businesses come up around a technology.”

If it’s hard to imagine AI as doing anything other than eliminating jobs, step back and consider the scope of the problem. The 2016 World Economic Forum report, “The Future of Jobs,” noted that “upcoming disruptions to the employment landscape are going to be a lot more complex and multifaceted than conveyed by a narrow focus only on automation”8 — saying, in a nutshell, that digital technologies and AI are not the only forces transforming the nature of work. It has been clear for some time that technological change — not just AI — obliges employees to become lifelong learners and embrace career flexibility, but as the WEF report observes, even more factors are at play: “technological, socioeconomic, geopolitical, and demographic developments and the interactions between them will generate new categories of jobs and occupations while partly or wholly displacing others. They will change the skill sets required in both old and new occupations in most industries and transform how and where people work.”9 Yet we have also seen digital technologies be used to address this problem. Accompanying the expansion of AI are many new learning options for humans: Augmented reality, new training tools, and digitally accessible forms of education (such as massive open online courses [MOOCs] and “nanodegrees”) are proliferating.

Against a canvas of even broader social, demographic, environmental, and global political developments, predictions of aggregate employment levels based on AI alone are difficult; there are too many countervailing forces to discuss any one of them in isolation. But it is not unreasonable to imagine an opportunity for AI to cushion some of its own impacts, and perhaps the impacts of other factors, by helping to anticipate the coming changes in the job market and identify (and meet) workforce training needs as they arise.

Even So, Inertia Is Not an Option. Big global uncertainties should not deter corporations from acting today, when action is required. Infosys, for example, has trained more than 120,000 employees in design thinking. This new capability will enable its employees both to shape a world of new AI-based service offerings and automate historic business processing services.

About the Research

To understand the challenges and opportunities associated with the use of artificial intelligence, MIT Sloan Management Review, in collaboration with The Boston Consulting Group, conducted its inaugural annual survey of more than 3,000 business executives, managers, and analysts from organizations around the world.

The survey, conducted in the spring of 2017, captured insights from individuals in 112 countries and 21 industries, from organizations of various sizes. More than two-thirds of the respondents were from outside of the United States. The sample was drawn from a number of sources, including MIT Sloan Management Review readers, and other interested parties 

In addition to our survey results, we interviewed business executives from a number of industries and academia to understand the practical issues facing organizations today. Their insights contributed to a richer understanding of the data.

For the purpose of our survey, we used the definition of artificial intelligence from the Oxford Dictionary: “AI is the theory and development of computer systems able to perform tasks normally requiring human intelligence, such as visual perception, speech recognition, decision-making, and translation between languages.” However, AI is evolving rapidly, as is the understanding and definition of the term.

Go to page No. 1  2 

No comments:

Post a Comment