Introduction

The number of sophisticated cognitive technologies that can cut into the need for human labor is expanding rapidly. But the key value of artificial intelligence isn’t to replace humans; it’s to sift through mountains of data and identify patterns — and problems — in real time. This collection of articles from MIT Sloan Management Review looks at the current applications and untapped potential of artificial intelligence.

From “When People Don’t Trust Algorithms”:

• We have a hard time letting go of human judgment — to our own detriment.
• Berkeley Dietvorst, an assistant professor of marketing at the University of Chicago Booth School of Business, says in a Q&A that even when faced with evidence that an algorithm will deliver better results than human judgment, most people consistently choose to follow their own minds.
• Dietvorst says that predicting student performance during an admission review is one example of an area in which an algorithm can outperform human judgement. Explaining that behavioral decision research expert Robyn Dawes “built a simple model” that can do just that, he says, “Take four or five variables — GPA, test scores, etc. — assign them equal weight, average them on a numerical scale, and use that result as your prediction of how students will rank against each other in actual performance. That model — which doesn’t even try to determine the relative value of the different variables — significantly outperforms admissions experts in predicting a student’s performance.”
• “The literature shows that, on average, when predicting human behavior, algorithms are about 10% to 15% better than humans,” he says. “But humans are very bad at it. Algorithms are significantly better but nowhere near perfection.”

From “Reshaping Business With Artificial Intelligence: Closing the Gap Between Ambition and Action”:

• The 2017 Artificial Intelligence Global Executive Study and Research Project by MIT Sloan Management Review in collaboration with The Boston Consulting Group presents a baseline that allows companies to compare their AI ambitions and efforts.
• The gap between ambition and execution is large at most companies. Almost 85% of executives believe AI will allow their companies to obtain or sustain a competitive advantage, but only about one in five companies has incorporated AI in some offerings or processes — and only one in 20 companies has extensively incorporated AI in offerings or processes.
• Combining responses to survey questions around AI understanding and adoption, four distinct organizational maturity clusters emerged: Pioneers (19% of organizations), Investigators (32%), Experimenters (13%), and Passives (36%).
• Expectations for AI run high across industries, company sizes, and geographic locations. While most executives have not yet seen substantial effects from AI, they clearly expect to in the next five years.

From “The Fatal Flaw of AI Implementation”:

• Many executives are enthusiastic about the business potential of machine learning applications. But business leaders often overlook a key issue: To fully unlock the benefits of artificial intelligence, they will need to upgrade people’s skills — and build an empowered, AI-savvy workforce.
• As with enterprise systems, AI inserted into businesses drives value by improving processes through automation. But the outputs of most automated processes also require people to do something, and processing isn’t worth much if people are feeding bad data into systems or don’t know what to do with analysis once it’s provided.
• Recruiting data scientists is not the biggest challenge. Organizations also need domain experts to help train AI systems to recognize important patterns and understand new data; they also need people who can use probabilistic output to guide actions that make the company more effective.

From “What You Need to Know Before Starting a Platform Business”:

• There has probably never been a better time for platform businesses. But that doesn’t make them easy to launch successfully.
• In a Q&A, economists Richard Schmalensee and David S. Evans, authors of Matchmakers: The New Economics of Multisided Platforms, say that what has created massive opportunities for platform businesses is the fact that, in effect, the internet is available everywhere. “Between mobile phones that people carry around, and then the internet of things connected to wireless networks, you basically have plugged the physical world into the internet,” says Evans.
• But building a platform has a chicken-and-egg problem: “Once you get lots of people participating on a platform, it can be really attractive to new participants, because there are a lot of people already there who they want to deal with,” explains Schmalensee. “But if you can’t get momentum, you don’t ever have an attractive product, because you’re not selling access to an attractive group of potential partners.”
• “Our perspective is that there’s too much hype about this business model,” says Schmalensee. “It’s really easy to look at companies like Uber and say, ‘Look at these wildly successful businesses!’ but forget all the people who tried to do similar things and failed.”

From “The Subtle Sources of Sampling Bias Hiding in Your Data”:

• Plummeting data acquisition costs have contributed to a surge in business analytics. But more data doesn’t inherently remove sampling bias — and in some cases, it could make it worse.
• Managers must be careful to understand how data was generated and how that might influence its value. The sources of bias in data sets can be subtle.
• Four practices can help: understanding the history behind your data; acknowledging that more data may not mean new data; recognizing that old data sources were imperfect, too; and remembering that intuition remains important.

From “What to Expect From Artificial Intelligence”:

• To understand how advances in artificial intelligence are likely to change the workplace — and the work of managers — you need to know where AI delivers the most value. Another way to put it: How does the technology reduce costs?
• The task that AI makes inexpensive and ubiquitous is prediction — the ability to take information you have and generate information you didn’t previously have. AI can help solve problems that were not previously prediction-oriented.
• Major advances in prediction may facilitate the automation of entire tasks. This will require machines with the ability to not only generate reliable predictions, but also use those predictions to determine what to do next. For example, for many business-related language translation tasks, the role of human judgment will become limited as prediction-driven translation improves (though judgment might still be important when translations are part of complex negotiations).
• The managerial challenge is threefold: Prediction is not the same as automation, the most valuable workforce skills involve judgment, and managing may require a new set of talents and expertise.

From “Digital Today, Cognitive Tomorrow”:

• Within the next five years, how will technology change the practice of management in a way we have not yet witnessed? “Digital is not the destination,” writes IBM Corp. CEO Ginni Rometty.
• “Rather, it is laying the foundation for a much more profound transformation to come,” she continues. “Within five years, I believe all major business decisions will be enhanced by cognitive technologies.”
• Most of the data being generated throughout the world today is “unstructured,” Rometty notes — including video and audio files, sensor outputs, and everything encoded in language, from medical journals to tweets. “Such unstructured data are ‘dark’ to traditional computer systems,” she writes, and therefore we need cognitive technology to process and use it.

From “Predicting a Future Where the Future Is Routinely Predicted”:

• Artificial intelligence systems will be able to provide real-time insights about business operations — as well as detect looming problems before they occur.
• The hunch-based bets of the past already are giving way to far more reliable data-informed decisions. But AI will take this further. By analyzing new types of data, including real-time video, AI systems will be able to provide managers with guidance about what is happening in their businesses at any moment in time. They will also be able to identify irregularities and issue warnings about problems that have yet to materialize.
• With AI, we can have machines look for millions of worrying patterns in the time it would take a human to consider just one. But that capability includes a terrible dilemma: if you check millions of things per hour, then you will receive hundreds of alerts every minute. In an effort to reduce false alarms, statisticians and AI researchers are working together to identify situations and conditions that tend to raise red flags that might not be warranted.

From “Just How Smart Are Smart Machines?”:

• Most managers don’t expect to see machines displacing knowledge workers anytime soon. Instead, they expect computing technology to augment rather than replace the work of humans.
• In the face of a sprawling and fast-evolving set of opportunities, the challenge is figuring out what forms the augmentation should take. Given the kinds of work managers oversee, what cognitive technologies should they be applying now, monitoring closely, or helping to build?
• To help, the authors developed a simple framework that plots cognitive technologies along two dimensions. It is illustrated in a chart titled “What Today’s Cognitive Technologies Can — and Can’t — Do.”
• Hardware and software will continue to get better, but rather than waiting for next-generation options, managers should be introducing cognitive technologies to workplaces now and discovering their human-augmenting value.