Читать книгу The AI-Powered Enterprise - Seth Earley - Страница 8

CHAPTER 1

THE PROMISE AND THE CHALLENGE OF AI

How will the future be different as a result of artificial intelligence? And what must your company do to stake its claim on that future? These questions are on the minds of every forward-thinking business leader, and they should be on your mind, too, whatever industry you are in. This book aims to provide you with answers to these questions so you can stop wondering what to do and get busy doing it.

When the web came along in the mid-90s, it transformed the behavior of customers and remade whole industries. New powerhouses like Amazon and Google arose and challenged established media and bricks-and-mortar retail companies. Even as Tower Records, Borders, and dozens of newspapers succumbed to the pressures of the digital revolution and the internet marketplace, other major players of that era—such as Home Depot and The Wall Street Journal—learned to thrive by capitalizing on the web’s vast new marketing reach and economies of scale.

AI has been called the greatest invention in human history—greater than fire, greater than the Industrial Revolution.¹ That’s hyperbole, but the magnitude of the change is indeed vast. AI is “reinventing the way we invent”² and will usher in a new era of human capabilities.

We are at an inflection point in human history.

That’s all well and good, you say. AI will change everything really soon—I think—but meanwhile my customers want lower prices while venture-funded startups are trying to steal those customers from me, my call center agents are overloaded with support calls, my competition is getting more challenging . . . and it takes my organization weeks to onboard new products. Meanwhile, inside the company, my employees continually complain that they “can’t find their stuff.” How are we going to take advantage of all of this great new technology if they can’t locate the marketing plan that worked so well last quarter or they have to recreate procedure documents yet again because they don’t know which ones are up to date?

The question for realizing the promise of AI is, How do you get there from here? At the 30,000-foot level, there are books explaining how AI will solve world hunger, while others are saying it will throw half of all humans out of work. At the other end of the spectrum, there are technical books that require a PhD in data engineering to understand. My intention is to be more helpful. This book will help you address practical issues and show you how to make the incremental changes that will prepare your company to deliver AI’s promise of revolutionary change. It will help you take the necessary steps down the path to transforming how your customers relate to your company. I will help you avoid career-limiting errors and I will show you how to set your BS detector for AI claims to “high.” I will help you catch this wave, rather than be crushed by it. This book contains a number of practical approaches for solving information problems. Without them, your AI will not work the way it needs to for your organization.

To understand what it will feel like to be a customer in this new world, let’s look in on a day in the life of one of those customers, an executive of a hypothetical manufacturer using hypothetical products from well-known brands, just a few years from now. (While Merrill Lynch, Grainger, and Expedia are real companies and may be working on chatbots, the bots in this scenario are invented representations of a fictional possible future.)

LIFE IN AN AI-POWERED WORLD

Allen Perkins feels powerful.

It is January of 2024. Perkins is the senior manufacturing manager for Hecker Heavy Locomotives, an Ohio company with a growing reputation for making the world’s highest-quality passenger train locomotives. Perkins feels powerful not because of the massive machines he helps to build, but because of the web of information that accommodates his every decision, whether at work or at home. His adept management of artificially intelligent resources has earned him a promotion, and now he must deliver.

As his self-driving Tesla heads for his office, Perkins can use the time to focus on his personal finances. The markets have been volatile, and he wonders if he should make any changes in his portfolio. He taps a button on the Tesla’s center console and speaks.

“Get me to Meryl,” he says.

“Meryl here,” a woman’s voice answers. Meryl is the bot for Bank of America/ Merrill Lynch, where Perkins’s investments reside. “What can I do for you today, Allen?”

“I’m a little nervous about the choppiness in the stock market. How am I doing these days?”

“Your net worth with us has risen by 2% in the last twelve months,” Meryl answers. “While the market has been going up and down by an average of half a percent a day for the last few weeks, we’ve got you in a pretty conservative mix of investments. So the value of your investments hasn’t been moving quite as much.”

“Are we on track with the kids’ college funds?”

“The investments for Ray and Sarah are likely to be sufficient to cover their tuition when they start college a few years from now,” Meryl replies. “I’ve been shifting those assets into more conservative positions because they’ll be in college so soon.”

“Is there a lot of cash in my account?”

“We’re about 10% in cash right now, Allen.”

“Can we invest some of that?”

“Sure, we could. But you’re going to need cash for that remodeling project you told me about last month. There’s not a whole lot lying around in your checking account right now. And don’t forget you just bought a bunch of new suits to go along with your promotion.”

“OK, leave the cash where it is,” Perkins says. “But let’s check back on this in a month or so. I may be getting a bonus.”

“Sure thing, Allen,” Meryl says. “Talk to you in February.”

Perkins’s Tesla pulls into a parking space near the front door and he strides into his office. After checking his emails, he begins the most important work of the day: selecting a new component supplier for six new locomotives for the California High-Speed Rail project.

He opens a browser and navigates to GraingerBot, a chatbot available from Grainger, one of the biggest suppliers to manufacturing companies like Hecker. He starts typing.

Hello, GraingerBot.

Hello, Mr. Perkins.

Let’s continue ordering parts for the California locomotives.

I have opened the file that we were working on yesterday. What would you like to work on?

Suspensions. Can you locate the suspension springs in the 3-D model?

I see eight helical springs that are 36.3 centimeters high.

What suppliers make springs like that?

Several make the chromium-vanadium steel specified in the drawings. But I’ve found some research that shows a new material may allow you to get more durability for the same cost. Shall I show it to you?

Please.

The GraingerBot pops up a twelve-page document from the International Journal of Engineering Research. Perkins reviews it and confirms that the GraingerBot is correct—this new material appears to be superior for this part. He resumes the typed conversation.

How many manufacturers can make this part from the new material?

Seven. That includes one parts supplier that you’ve already used for six other components; they’ll give you a discount if your order pushes the total over $35,000.

Get quotes on those, GraingerBot. And get a few quotes on the chromium-vanadium springs, too, just in case we decide to stick with them.

I’ll have those answers for you by 12:30. I’m connecting with the bots for those parts suppliers right now.

Perkins signs off. This has been productive. While he would have liked to complete the order for the springs now, it looks like he might have found a more durable material—and with Hecker now doing the maintenance as well as the manufacturing, that would save the company some serious cash.

Perkins turns his attention to an upcoming business trip. He is going to Sacramento to meet with the California transit team that is buying these locomotives. He puts on his Bluetooth headset and connects with Pete, his virtual travel agent from Expedia.

“Pete, get me a flight to Sacramento for next Tuesday.”

“Sure thing, Mr. Perkins,” says Pete. “When will you be returning?”

“Wednesday or Thursday. Look into both.”

“OK. I’m checking American Airlines, because that’s where you have the best status. I have 18 possible flight combinations. I’m afraid you will have to connect.”

“Yeah, nobody flies nonstop Dayton to Sacramento! Don’t route me through Chicago or Denver. Too much snow this time of year,” Perkins says.

“That leaves nine possible flights. Is this for the two o’clock meeting with the California rail project team?”

Perkins remembers that he has given Pete access to his schedule, and realizes the bot is figuring out flights that fit his existing appointments.

“Yes, that’s the meeting,” he says.

“I’ll make sure your flight gets in well before that. I see that one of your manufacturing partners is based in that area. In fact, it looks like GraingerBot is connecting with them right now to check on some springs. Should I set up a meeting?”

“Definitely. See if you can get me in to see them on Wednesday. I’d love to tour that facility.”

“Do you want to take a red-eye back?”

“No way,” Perkins replies. “Better make it Thursday morning.”

“OK, I see the best roundtrip now. It leaves at 7 a.m. on Tuesday and your return flight takes off at 8 a.m. on Thursday. Will you be staying at the Hilton at the airport again? They’ve got the best price for a three-star hotel.”

“Go ahead and book it. And get me a rental car, too. Thanks, Pete.”

Perkins muses for a moment about travel reservations. All that clicking and checking he used to do. . . . But the bots pulled it together in a way he never could have before. Perkins likes the feeling of control that gives him—and the efficiency, too.

Perkins doesn’t need to think about what is going on behind the scenes to make investing, manufacturing, and travel so much easier than they used to be. But every time he interacts with a bot, he is drawn deeper into an ecosystem of artificial intelligence. He has grown to rely on the companies that were making his life easier—those that were ready with AI. Everybody else is working with those companies, too, and the competition is slowly fading into irrelevance.

WHAT IT TAKES TO SUCCEED IN THE AI FUTURE

We’re on the verge of the future that Allen Perkins lives in. But not every company will make it into that future.

We know how we want our companies to work. Enterprises ought to be customer-focused, responsive, and digital. They should deliver to each employee and customer exactly what they need, at the moment they need it. The data and technology to do this are available now.

Any big company is likely to have an abundance of technology. It has systems for customers, inventory, and products, along with websites and mobile apps. These systems are spitting out data all day long. Within that data is exactly the information needed to make a business more responsive. The problem is, the data is often not used as it could (and should) be.

Artificial intelligence ought to take that data and turn it into effective execution, just as it did for Allen Perkins. IBM’s Watson and Amazon’s Alexa seem pretty smart. But despite the billions of dollars spent so far on bots and other tools, AI continues to stumble. Why can’t it magically take all that data and make an enterprise run faster and better? Why can’t it deliver the experience that Allen Perkins would expect?

Our organizations are up to their eyeballs in technology, and every venture capitalist believes that yet another tool is what the industry needs. But even after multiple generations of investments and billions of dollars of digital transformations, organizations are still struggling with information overload, with providing excellent customer service, with reducing costs and improving efficiencies, with speeding the core processes that provide a competitive advantage. Why is this happening? Because the foundational principles that I will discuss throughout this book are ignored, given short shrift, deprived of resources, or considered an afterthought. The elements that are required to make the shiny new technologies live up to their promise require hard work that is not sexy and shiny. There are new tools and approaches that make these efforts more efficient, and ways to embed new approaches to dealing with information and data, but they still require discipline, focus, attention, and resources.

Perhaps your organization has experimented with AI. An executive at a major life insurance company recently told me, “Every one of our competitors and most of the organizations of our size in other industries have spent at least a few million dollars on failed AI initiatives.” In some cases, technology vendors have sold “aspirational capabilities”—functionality that was not yet in the current software. But in most cases, the cause of the failure was overestimation of what was truly “out-of-the-box” functionality, overly ambitious “moonshot” programs that were central to major digital transformation efforts but unattainable in practice, or existing organizational processes incompatible with new AI approaches. Leadership may have bought into the promise of AI without adequate support from the front lines of the business. Technology organizations may not have been adequately prepared to take on new tools and significant process changes. In many cases, the technology may have been potentially capable of functionality, but the data, locked in siloed systems, was inaccessible, poorly structured, or improperly curated.

Many AI programs attempt to deal with unstructured information and replicate how humans perform certain tasks, such as answering support questions or personalizing a customer experience. That may require pulling information from multiple systems and weaving together multiple processes, including some that have historically been done manually. To deliver on its promise, AI needs the correct “training data,” including content, metadata (descriptions of data), and operational knowledge. If that data and corresponding outcomes are not available in a way that the system can process, then the AI will fail.

How do you make those data and outcomes accessible to power the AI? That’s where the ontology comes in.

THE CENTRAL ROLE OF THE ONTOLOGY

AI cannot start with a blank page. It leverages information structures and architecture. Artificial intelligence begins with information architecture. In other words, there is no AI without IA.³

AI works only when it understands the soul of your business. It needs the key that unlocks that understanding. That’s the science behind the magic of AI. The key that unlocks that understanding is an ontology: a representation of what matters within the company and makes it unique, including products and services, solutions and processes, organizational structures, protocols, customer characteristics, manufacturing methods, knowledge, content and data of all types. It’s a concept that, correctly built, managed, and applied, makes the difference between the promise of AI and delivering sustainably on that promise.

Simply put, an ontology reveals what is going on inside your business—it’s the DNA of the enterprise.

An ontology is a consistent representation of data and data relationships that can inform and power AI technologies. In different contexts, it can include or become expressed as any of the following: a data model, a content model, an information model, a data/ content/information architecture, master data, or metadata. But an ontology is more than each of these things in themselves. However you describe it, the ontology is essential to and at the heart of AI-driven technologies. To be clear, an ontology is not a single, static thing; it is never complete, and it changes as the organization changes and as it is applied throughout the enterprise.

The ontology is the master knowledge scaffolding of the organization. Multiple data and architectural components are created from that scaffolding, so without a thoughtful and consistent approach to developing, applying, and evolving the ontology, progress in moving toward AI-driven transformation will be slow, costly, and less effective. The components of the ontology are the ones we have mentioned: metadata structures, reference data, taxonomies, controlled vocabularies, thesaurus structures, lexicons, dictionaries, and master data correctly designed into the information technology ecosystem. The ontology is at the heart of the information design of the AI-powered enterprise and it becomes an asset of ever-increasing value.

While it is true that some algorithms can operate on data without an external structure, they still operate based on the features programmed into the underlying system. Even if there is no structure to the raw data, the algorithm will perform better if more of that structure is provided as an input—as an element of the ontology.

Ontologies are a complex topic. For that reason, I’ve dedicated a whole chapter to them: chapter 2. For now, just know that the ontology is what makes the difference in whether AI drives your enterprise forward or just adds to the incompatible welter of technology that is slowing you down.

ORGANIZATIONS ARE LIKE BIOLOGICAL ORGANISMS

I like the economist Gareth Morgan’s metaphor likening businesses to organisms living in an ecosystem competing for resources.⁴ It explains so much about why AI projects go wrong.

An organism’s survival depends on (a) perception of information from the environment (b) correct interpretation and processing of that information, and (c) communication of information as signals that are sent quickly to the parts of the organism that “need to know” and act. This process requires efficient internal communications and coordination so that resources can be deployed in response to the signal for a swift and appropriate response.

Just like organisms in an ecosystem, businesses consume energy and resources and then create solutions and structures from those resources. The resources and results primarily take the form of information: businesses are living organisms that consume and produce information. Their agility and adaptability depend on how effectively they metabolize that information.

For example, consider how our brains and bodies act on signals from the environment and interact with the world based on integrated information systems and feedback loops. When the amygdala (the part of the brain that registers fear or desire) identifies a threat, our sympathetic nervous system (which controls the “fight or flight” response) reacts in a highly orchestrated way. Another part of the brain—the hypothalamus—instantly sends a signal throughout the body. This triggers the adrenal glands to release adrenaline, which causes a cascade of responses that we are all familiar with from instances when we are startled, such as if a car speeds toward us as we step into a crosswalk. The heartbeat increases, breathing becomes more rapid, and we feel a surge of energy. The brain also executes a new computational task—coming up with the appropriate expletives to hurl at the driver—and anticipates likely outcomes (Yikes, is he getting out of his car?). Everything works holistically to respond efficiently and effectively to the stimulus, with very little friction.

It’s easy to see why holistic and synchronized information flows are essential to survival. It would not do us much good if the brain had to rummage around our past memories of speeding cars and try to decide what to do. The same kind of holistic, synergistic, and simultaneously integrated flow of information is also what’s needed to create transformative AI solutions of the sort we read about in Allen Perkins’s story.

For an organization to function effectively, the systems for managing information and the processes for supporting information flows have to be flexible, adaptable, and responsive to market conditions. This is the “organic” nature of the organization.

In this organic view of the enterprise, the subsystems within companies, such as business units and departments, are analogous to the organs and biological systems in an organism. For these departments and functions to operate smoothly, their information needs have to be met. Marketing needs product specifications and features from the enterprise’s engineering and product development functions. Sales requires support collateral and leads from marketing. Finance needs customer and order information from sales. Each of these functions needs reliable information to achieve its objectives.

Let’s take the metaphor further. Why does it matter how effectively information flows in the company? Because the entire economy is a collection of organizations interacting in an ecosystem that follows the same principles that living biological systems follow. Just as in a biological ecosystem, the economy is a collection of entities operating in interdependent networks and competing for resources. In the case of business, those resources include time, money, talent, expertise, and the attention that ultimately comes from customers. Information flows—both within companies and in the broader economy—are at the center of this activity. Our society and its associated value chains, knowledge networks, and information flows are endlessly complex, interrelated, and nuanced, containing layers of systems upon systems at every level of interaction—from the most basic to the most mind-bogglingly complex.

Just as environments select for the strongest organisms in a particular set of conditions, so too does the economy select for the strongest and best-adapted organization in a particular market sector or niche. Many companies will go extinct as our economic environment continues to change. Yours can be one of the survivors, but only if it can adapt to those changing conditions.

SERVING CUSTOMERS EFFECTIVELY IS WHAT MAKES COMPANIES SUCCESSFUL

In biological systems, organisms succeed because they can find food and reproduce (grow). What is the analogy to this in business? It is serving customers. The enterprise that best serves the customer grows and thrives at the expensive of its competitors.

While customer-centricity is ostensibly a major focus of most enterprises today, in many cases the understanding of customer experience is immature, or is siloed in parts of the organization that “own” one aspect of customer service or customer support. That understanding of customers is largely disconnected from other critical processes. Using the entire customer journey as the anchor to programs—whether traditional information projects or cutting-edge machine learning programs—is critical to success. (I describe the customer journey in more detail in chapter 3.) The customer journey is not a single thing; it encompasses everything that the enterprise does, directly and indirectly, that impacts how its customers perceive their interactions and the value they receive. If a system or process cannot be tied to value for the customer (whether that customer is an internal consumer of information or an external paying-end customer), then it cannot be justified.

The real challenge to delivering value occurs when systems are part of the foundation for a capability that can impact the customer, but customers don’t interact directly with those systems. Infrastructure programs are notoriously difficult to justify because of the lack of a clear line of sight to their impact on the customer; but when the infrastructure is faulty, serving the customer becomes slow, complicated, or impossible. When the customer is internal, this is even more of a challenge. (I will discuss the “internal” customer in chapter 8 on productivity, but the same principles largely apply in that context.) In chapter 3, I will define an approach and framework that will show you how to establish the linkage between systems and customer value and how to communicate the value any project in the enterprise has for the customer experience.

Some of these linkages are indirect. Others may be difficult to measure. However, they all should be traceable to activities and capabilities that produce customer value. Everything that an organization does needs to improve how it serves its customers and how the enterprise creates value; each program, project, investment, and decision should be linked to a metric that impacts the customer’s perceptions of their experience and the value that they receive from their relationship with your organization. High-fidelity customer journeys (described in chapter 3) are the key to successful AI initiatives such as personalization and campaign optimization, which can transform the customer experience and deliver new efficiencies and capabilities from human augmentation and machine-enabled automation.

AGILITY AND ADAPTABILITY DETERMINE COMPETITIVE SUCCESS

It’s not enough to be customer-centric. You must also operate quickly. Speed matters in the biological world (ask the early bird that gets the worm), and increasingly, speed and agility are the factors that differentiate successful enterprises from those that lose the race.

Every enterprise needs to adapt to changing conditions in the environment (including customer needs, competition, changes in technology, and macroeconomic shifts). The faster it can do so, the better it can serve customers and beat the competition. Getting products and solutions to market faster requires faster decision-making, faster feedback, faster iterations, and faster experimentation—and each of these requires faster information flows. Friction in processes, systems, and technologies impede the flow of information and therefore slows decision-making.

Unfortunately, the speed of business change is inherently faster than the speed at which complex enterprise systems can adapt. When systems have data in incompatible formats, or when that data is of poor quality due to manual processes and conversions, making the data compatible adds another layer of friction. Friction comes from many sources, including out-of-date information, lost content, incorrect or poor quality data, disconnected processes, incorrect measures, inconsistent communications, excessive meetings, overcommunication, and undercommunication. These sources of friction become impediments that inexorably slow information flows.

The challenge is that this happens in myriad incremental ways that are difficult to pinpoint, hard to quantify, and challenging to remediate. In most cases, this friction is accepted as business as usual or as the nature of the beast—“that’s just the way things work.” Or leaders decide that it is too expensive to fix and that adding some people to deal with the issue is the most economical (short-term) solution. The problem is that these incremental friction points, which individually do not make sense to address, are never considered as a whole—or if they are, they become too difficult and disruptive (not to mention costly) to tackle.

AI projects can provide incremental value by addressing these friction points. As Tom Davenport points out in his book The AI Advantage, AI and cognitive technologies augment human work and provide incremental value in many areas that accumulate to transform the enterprise.

Since organizations operate on information flows, the leadership of the enterprise needs to invest in the things that smooth, speed, and improve the efficiency of those flows. The problem is knowing what to invest in and how to harvest the benefits of investments. When executed correctly and based on a solid ontology, investments in data and configuration of the infrastructure through which data flows can lead to enormous value. One implementation that I will describe led to a $50 million annual savings from a $1 million investment. Another digital transformation, the core of which was built on investments in data, data architecture, and technology infrastructure, led to an $8 billion increase in market value from a $25 million investment in the foundational principles that I will discuss throughout this book.

Things are accelerating as new entrants arrive that are “born digital”—like Airbnb in the lodging business, Tesla in the automotive business, Google in media, or Uber in transportation. Such businesses have the luxury of reinventing the end-to-end value chain and building integrated processes that speed the flow of information and data using a green field/ clean sheet approach. It’s a huge competitive advantage—like the speed that warm-blooded animals tapped during the evolution process to allow them to compete effectively with sluggish reptiles.

THE PROBLEMS THAT KEEP BUSINESSES FROM MOVING FASTER

Businesses know all of this. So why can’t they get out of their own way? Let’s review the three most common problems that stop corporate organisms from competing effectively in the economic ecosystem.

Problem One: Friction and Siloed Communication

Companies, especially those born in the pre-digital age, are created not holistically but in a piecemeal fashion. Departments and functions operate independently. When a signal comes in, it doesn’t instantly go everywhere it needs to go; it must be routed from one place to the next. Internal communications, databases, and other systems act as messengers to relay a signal throughout the organization. Imagine if the brain told the legs to run, then the legs had to go and ask the adrenals for some energy to execute the movement, and the adrenals had to then go and verify that the decision was authorized by the brain. Before you could move, you’d be run over.

In most companies, there is a lot of friction along the way. Challenges like differing or competing priorities, incomplete understanding of the big picture, dropping the ball, and double-checking all create drag as information makes its way through the organization. That’s not fatal when the information being handled is human brain–sized, which is why manual systems and processes have worked for companies so far. But it collapses under the weight of the enormous data sets required for AI and high-tech interventions and for the rapid response required in a digital economy. The key to making AI work is reducing that friction and speeding the flow of information.

Problem Two: Incompatible Data and Language

From time to time, every organization needs to do housekeeping on its digital working files, archived data, and other forms of information. But how do you do that consistently and cleanly when every process or function uses different terminology or a different way of organizing things? Too often, the various groups within an organization are speaking their own languages—or at least their own dialects. Information is stored differently, and teams use different terminology that speaks to their needs and processes but that doesn’t consider the broader organization. This leads to manual workarounds and to information getting lost in translation. As I’ll show in more detail in chapter 8, technology companies designed collaboration tools to make it easy for people to create and use data, not to make the data effective in the context of a corporation. The result is an information environment full of poorly designed, fragmented, and disconnected systems. It’s not a surprise that many of these systems don’t share a consistent set of data language.

Problem Three: Junk Data

Entropy is another fact of the physical world mirrored by the digital one. Every system tends from order to disorder, and reestablishing order requires energy. We all experience this in our day-to-day lives: our desk gets messy and we need to put energy into organizing it. The house gets dirty and requires energy to clean. Information gets messy, too, and organizing it also takes energy. For example, we have to delete or label our email messages, put files into folders, or cleanse a document repository of any out-of-date material. Productive activity seeks to reduce the amount of disorder and therefore reverse the entropy of a local system through the application of energy.

Most organizations today are drowning in junk data as the incredible volume of digital information produced massively outstrips the energy available to manually practice good data hygiene habits. This book will show you how to get your digital house in order so the robots can keep it clean for you.

TAGGING UNCLOGS THE FLOW OF INFORMATION

Tagging is a central part of what makes ontologies able to speed the performance of enterprises.

A manager’s objective is to give employees direction; provide resources and the information necessary to solve problems; and allow creativity, hard work, and expertise to generate solutions that have value to customers and the marketplace. All of this activity is fueled by knowledge, and the way that knowledge flows through the organization’s networks is key to its efficiency and effectiveness. Important information needs to be flagged, tagged, and held up as meaningful. This information includes, for example, the needs of customer segments based on market research, solutions to engineering problems, the current quarter’s strategic objectives, and the features of a new product and how it is different from the competition. These are all signals that need to be separated from the noise of day-to-day communications.

The separation comes from tagging that identifies the information as important. Then someone can take that important piece of data and use it to solve their problem. (As I will describe later, that tagging, or separation of signal from noise, can happen at multiple levels—from manual information and data curation performed by humans through AI and machine learning approaches.)

Not having the right tags causes meaning to be lost—the noise drowns out the signal. Inefficiencies in information flows, lack of consistent terminology, and systems that don’t talk to one another bog down operations and create waste. We’ve all seen what this looks like and felt the pain: for example, folder structures on shared drives that are redundant or nonsensical, or that contain labels meaningful only to individuals (“Joe—Important docs”), along with excessive translation and manual manipulation of data because systems do not use the same terminology or data standards.

Constraints can be liberating. The enterprise adapts and thrives when simple tagging rules that speed information flows create value through emergent behaviors within the system. Artificial intelligence is a mechanism that, properly designed and applied, speeds information flow and enables those emergent efficiencies, including tagging. It can therefore help every part of your organization do its job more quickly, efficiently, and consistently. To succeed over the long term, you must allocate resources in a sustainable way to build AI-ready assets of increasing value—including the ontologies and data structures that will serve all aspects of the business. The business needs to use continuous feedback to develop these structures, apply them to the production of value, fine-tune systems and processes, and adjust and make course corrections.

What does such a solution look like? There are three areas in which your company needs to rethink its resources, focus, and attention so it can exploit the power of artificial intelligence. My systematic plan that you can follow to prepare for an AI-powered enterprise includes these elements:

1.Data. This includes how data is architected, managed, curated, and applied. You must wrangle your messy and inconsistent data into a refined asset for high-precision, high-leverage activities. The organizations with the most agile architecture, highest-quality data, and best algorithms for applying that data to address customer and employee needs will win.

2.Technology. To deliver personalized experiences to customers (whether internal or external), appropriate technologies must scale processes by relying on detailed enterprise knowledge. They must also remain adaptable as tools, approaches, and information sources evolve and change.

3.Operationalization. Just as with the reengineering transformation of the ’80s, these improvements require a commitment to new forms of organizational discipline, with new accountabilities and metrics. This includes rethinking how the organization delivers value through end-to-end digital processes.

Let’s examine each of these three requirements in detail.

Data: The DNA of the Organization

Most executives reflexively nod their heads when discussing the value of data. “Data is extremely valuable to our enterprise,” they say. “We are undergoing a data-driven digital transformation.” According to this thinking, good data is good, and bad data is not good. But when it comes to creating ontologies—which means fixing the foundational data issues in a sustainable way—and they see the price that the enterprise needs to pay, these same executives deprioritize projects like fixing the data-quality issues. “It’s not that valuable” is the message that is communicated to the organization. Because data issues are not addressed properly at a fundamental level, tens or hundreds of millions of dollars will be spent on digital transformations that will fail in the long run.

High quality, findable, and usable data is an essential part of the AI-powered enterprise. Machine learning can find patterns in unstructured data, but it cannot make sense of information that is of poor quality or missing. Data needs to be contextualized; you cannot simply “point the AI at all of your data” as some in the industry have claimed. Depending on the application it can also require curation and structuring for ingestion into many classes of AI tools, including cognitive systems. More structure will allow the algorithm to function more precisely.

For cognitive applications such as chatbots, the required training data is the same information that humans need but with a different structure and format. Predictive modeling AI needs training data to build recognition patterns and examples to learn from. Data is more important than the algorithms, and bad data will provide bad results.

Perhaps technology vendors have assured you that their AI will fix your data. That’s optimistic.⁵ In fact, many of the AI technologies on the market are actually Band-Aid solutions that try to make up for our sins in data and content curation. Because of a lack of resourcing and poor data hygiene, organizations are paying the price, and that price includes trying to fix data with AI, even though it’s the organization’s own data processes and governance that are at fault. Yes, AI can help, but there’s more to it than what the large systems integrators are telling you.

The way to fix this problem is to harmonize your data with consistent data structures and models, creating a Rosetta Stone that helps your systems communicate and provides a waypoint so your AI can navigate your messy, fast moving, diverse, unstructured and structured data universe. That Rosetta Stone is the ontology.

While many master data types are challenging and prone to failure, the approaches we will discuss will increase your chances of success and will move the needle on multiple projects. Rather than striving for a “single source of truth,” the goal is to increase the consistency and quality of information so that there is less friction throughout the organization. There will be differences in organizing principles and structures, but rather than being accidental, they will be intentional. The ontology becomes a reference point to inform where information structures need to be harmonized and where there can be (intentional) differences.

Technology: Having the Right Tools to Serve Customers (Internal and External)

Choosing the right technologies and integrating them successfully is a critical part of building out your AI program. It is not simply about adopting machine learning tools and technologies that are labeled as “AI,” because, these days, every technology vendor says what it does is “AI.”

Forget AI for a moment. Recognize that knowledge workers of all types, from engineers to marketers to designers, interact with technologies to accomplish their day-to-day tasks that directly or indirectly support the ways that customers interact. These workers work with systems on their smartphones, laptops, tablets, and intelligent connected appliances and devices. Having the correct technologies integrated in an adaptable, flexible way allows for processes and functionality to evolve in parallel with customer needs and the competitive landscape. Having the right tools enabled and enhanced by AI to serve each stage of the customer journey makes the business run faster, better, and in a way more aligned with customer needs.

The real challenge is that, if you’re like most organizations, you probably have a patchwork of systems, technologies, and processes that have evolved organically over time. These have inevitably led to messy and complex integrations, manual processes, and workarounds that make things more difficult in the long run. Some people refer to this as technical debt—the shortcuts taken in the hope of deploying technology more quickly. But just like debt in the real world, this approach is costly, and you have to pay for the shortcuts, sometimes at interest rates that would be classified as usury if they were on a consumer’s credit card. What appears to save time and money in the short run, when accumulated over numerous projects and multiple years, hamstrings the business and prevents adaptation as each leader kicks the can down the road for the next person to deal with. And now that person is you.

Adapting legacy systems built on a patchwork of different platforms and technologies to changing conditions is a slow and laborious process. But it’s essential to making the tools and technology effective and the data actionable.

Operationalization: Leading with Vision and Managing Change

Every project, whether a departmental reorganization or a ten-year growth plan, begins with vision and strategy. What is different about an AI strategy is that the possibilities are entirely new, which means you will have to look at the business and customer relationships in entirely new ways. Whenever there are fundamental shifts in technologies, what becomes possible is not just an extension of where we are but an entirely new way of being or interacting. If you don’t know what is possible, you cannot think about those possibilities and consider what they mean for the business.

In the early days of the web, organizations were just thinking about how they might use it to get their marketing materials in front of more consumers. They were not thinking about the capabilities of iPhones to carry the equivalent of shelves of compact discs or every photo album owned, or about what it might be like for consumers to carry a digital bank teller in their pocket. They certainly weren’t imagining entirely new business models like ridesharing services. Those capabilities evolved and required the concurrent evolution of various supporting components. Some organizations were ahead of the curve and disrupted industries, while others underinvested and were left behind. Still others overinvested before the market was ready and wasted resources.

The other operational challenge is that the different parts of the organization—internal systems and processes—move at different speeds and change at different rates. For example, enterprise resource planning (ERP) systems are relatively stable and do not change frequently. A long development cycle is required to add or evolve core modules and functionality. At the other end of the spectrum are social media applications and programs that change extremely rapidly. Inherent mismatches in these clock speeds in the organization exacerbate the data and architecture challenges (see Figure 1-1). As the AI landscape evolves, governance structures, processes, and decision-making need to allow for adaptability and to support cultural change that will be part of the new organization dynamics.

The challenge is that business always changes at a faster pace than the internal IT organization can support, and technology changes faster than humans and organizational processes can absorb. But even if IT could keep up with the best-of-breed tools for the business, people would not have the capacity to absorb change that quickly.

Figure 1-1: Varying Clock Speeds throughout the Organization

Decisions about the pace of change need to be methodical and data driven, not based solely on opinions. A metrics-driven framework for managing decision-making and resource allocation removes guesswork and ensures that your investments will produce value. Sustainable, metrics-driven governance is the single determinant of successful AI programs. Throughout this book, and especially in chapter 10, I’ll show you how to set up a governance playbook that can be updated and evolve as the capabilities of your AI-powered enterprise continue to mature.

Governance is built around organizational structure and reporting models, aspects that need to evolve and be updated as capabilities mature in AI-powered organizations. For example, old school “spray and pray” mass marketing approaches and skill sets need to be updated with data-driven digital marketing precision. The workforce will have a new makeup in the AI-powered enterprise, and just as machinery amplified physical human strengths, machine intelligence will amplify cognitive human strengths.

In this chapter, we began to explore in broad brushstrokes how your organization needs to transform foundational processes and further evolve fundamental capabilities as you embark on your AI-powered path forward. Decisions in each of the three areas—data, technology, and operationalization—need to be grounded in the organization’s strategy for how it will serve the customer, and must be based on data-driven approaches, not trial and error.

Ontology Supports Everything

All of these elements—data, technology, and operationalization—relate to having consistent fundamental organizing principles reflected in the ontology. Ontologies capture more than data representations. These structures begin at the level of concepts that are important to the business and are then translated into processes, systems, navigational constructs, applications, and yes, data structures. In addition to allowing systems to talk to one another, these naming conventions, organizing constructs, conceptual relationships, and labels speed information flows; enable data, content, and information to be integrated; streamline end-to-end processes; permit the aggregation of data sources; and stimulate faster adaptability in a rapidly changing ecosystem. They allow for the contextualization of insights, the cataloguing of metrics, and the creation of feedback mechanisms for continual improvement.

In the end, the foundation of an enterprise ontology captures the intelligence of the organization and becomes the scaffolding that guides and optimizes information flows, powers and contextualizes insights from AI systems, and becomes the brains behind tools like chatbots and cognitive search.

COGNITIVE AI AND ONTOLOGIES

Many of the examples in this book are from a group of applications referred to as “cognitive” AI. This class of AI seeks to improve how humans interact with computers and requires an intentional approach to building out the underlying knowledge architecture—or, as we discuss throughout the book, the ontology.

There are other classes of machine learning and analytics types of AI whose algorithms may not depend as extensively on this knowledge engineering approach. However, I want to make two distinctions here. While this book focuses a great deal on cognitive types of AI, I will show how other types of AI that leverage a range of purely data-centric approaches (deep learning and neural networks, for example) can also work better with defined data structures—including knowledge architectures and ontologies.

For example, machine learning algorithms may not need an ontology to function, but applying the results to the business does require the consistency and efficiency provided by an ontology and the resulting knowledge architecture. Similarly, while a neural network may not require an ontology to perform the analysis, application of that analysis does.

Many of the approaches in this book are part of a technology-agnostic tool kit for making changes and improvements that lead to greater efficiencies, increased revenue, and greater differentiation in the marketplace. In many organizations, these approaches have not reached their full potential with current technology. Many of the problems that organizations are turning to AI for are in part due to the fact that the correct approaches have not been operationalized using existing technology. In other words, we are using AI (or trying to do so) to make up for our past sins in poor information hygiene. This is the nature of the industry—fast-changing tools and the inability to absorb and manage change effectively over generations of technology adoption. These approaches are even more critical in today’s competitive landscape.

YOUR CHARGE FOR THE FUTURE

This is a practical book for CEOs, CMOs, and technology executives who want to transform their business to get a jump on the opportunities of the AI future. It’s not just about the technology of AI. It’s about how to manage the change, step by step. It’s also about understanding where to get the money and where the quick wins are. In short, it’s about learning what a business driven by ontology-powered artificial intelligence can do and how to make that happen.

Gaining this edge depends on establishing a foundation that includes pieces of the organizational and technology puzzle. Some of these are likely familiar—such as having the right people supported by the right tools and processes—but they will require new approaches, while others are entirely new (such as fresh ways of looking at how technology can support your customers). In some cases, the missing ingredients will be a new sense of discipline and an increased level of resourcing and commitment applied to known approaches to problems.

To get this edge, executives at your enterprise must create a vision that empowers AI to uniquely deliver your value proposition in support of your customers’ experience, as well as a strategic plan for differentiating from your competitors. It will require developing new supporting processes and new competencies. I will outline the steps for benchmarking the maturity of each of these components and developing a plan for bridging gaps between where you are and where you need to be.

No twenty-first-century executive can succeed without understanding the role of AI in enterprises and how to make this technology work effectively. It’s time to go beyond clouds, big data, and mobile fantasies. It’s time to learn what it takes to power your enterprise with AI, now.

TAKEAWAYS FROM CHAPTER 1

In this chapter, I’ve described how AI will change the priorities for enterprises and how ontology can play a critical role in taking advantage of those opportunities. These are the main points in this chapter:

•AI marks an inflection point in human history.

•AI initiatives fail if the information they depend on is not properly structured.

•The ontology—a central repository of data terms and relationships—is what makes it possible for AI projects to succeed.

•Organizations are like organisms that depend on freely flowing information to survive.

•Agility and adaptability are the qualities that enable organizations to thrive in their broader ecosystem.

•The failure of holistic communication, data incompatibility, and junk data prevent companies from operating effectively.

•Tagging speeds the flow of information.

•Competitive advantage arises from the ways that organizations manage data, technology, and operationalization.