We are on the verge of the artificial intelligence (AI) gold rush. Like the prospectors of the infamous historical gold rush, however, only a few leading organizations will strike gold.
Real economic growth will be achieved by the companies selling the equivalent of picks, food, supplies, shovels, and jeans for artificial intelligence and machine learning. Think of all the tools required: training data, governance tools, consulting and integration services, and most critical, the creation of new sustainable revenue models.
Startups, incumbent tech companies, and corporate innovation centers have already started using artificial intelligence and machine learning to solve real business problems across nearly every industry, including manufacturing, healthcare, transportation, and energy.
The first successes were no-nonsense: Take a process that you know well and move the heavy lifting to AI, enabling humans to do more creative thinking. This approach was characterized by short-term wins, intended to be cross-company scalable with a focus on immediate value creation.
“There are many parallels between Internet retailing in the nineties and artificial intelligence today: Those that embraced electronic intermediaries were able to redefine and grow within a click and mortar world, whereas others have seen their brick and products fade and seem digital substituted. In both situations, commoditization, or fading away, awaits those that wait too long to ride the wave of social disrupting technology.”
2020: Augment Existing Jobs and Transition to Jobs Yet to Be Created
Over the next two to five years, we can expect a profound transformation for knowledge workers and professionals as their daily tasks are infused by AI and machine learning. Unlike the stuff of science fiction movies, AI and machine learning will not do the job autonomously; AI will instead relieve the human from repetitive work, and force (or assist) humans to make decisions faster and easier. It is just the opposite of what we anticipated — humans must decide when to tell machine learning systems to do the work. This is a prime example of the human component of machine learning and the importance of creativity when machine learning is in play.
Leveraging technology, like using artificial intelligence in processes and augmenting tasks, will actually strengthen the economy. Take JFK’s premise as a truth: As we have the talent to invent new machines that reduce jobs, we also have the talent to invent new jobs. The AI gold rush will drive the creation of new AI jobs. Gartner predicts that by 2020, artificial intelligence will create more jobs than it eliminates. Roles like citizen data scientist, best practice training data creator, and AI trainer will be needed for a variety of industries and domains — think regulatory, law, or finance.
It will also create executive jobs such as the chief data officer, AI ethics and governance officer, or AI training-property protection — the secret sauce of how companies do things. There will also be data monetization-related jobs, where companies will see both monetization of their AI-enriched data and AI-trained data services to their industry or value chain.
As per Gartner’s prediction, the number of jobs affected by AI will vary by industry; through 2019, healthcare, education, and the public sector will see continuously growing job demand, while manufacturing will demonstrate the greatest growth. AI-related job creation is predicted to cross into positive territory starting in 2020, using AI where it matters, reaching two million net-new jobs by 2025.
The year 2020 will be a pivotal moment of mainstream AI usage. Various analysts have already projected that by 2020, around 70 percent of the data that a company uses will come from external data streams and IoT devices. By 2020, experts predict that 50 billion things will become connected to the Internet.
To put this in perspective, that means nearly seven connected things for every person on the planet. As billions of things, data, business processes, and people become connected, artificial intelligence, powered by machine learning, will need to do the filtering, inference, and prediction to make this all work. The current predictions are that $19 trillion of value will be created, justifying the term AI gold rush.
Going for Gold Versus Making Gold
The core question that companies need to ask relates to their core competence: “Is data the core asset that I monetize?” or “Is data the glue that connects the processes that have made my products or services successful?”
Let’s start with companies that are purely data-centric, monetizing their data for product selection and insight: For these companies, data is everything. Data is their asset. Data-centric organizations will have to create their own business model to find their gold.
Data-Centric Companies That Create AI
A few years ago, many people thought that the creative part of streaming video services, such as Netflix or Amazon Prime, would be marginal. The recent Emmys and international film festival awards have proven otherwise. As we choose our entertainment today, models, data science, and machine-learning assist our choice, predict our interest, and drive investment for new original content.
Preparing the data used to train your own baseline AI models is incredibly time- and labor-intensive. For data-centric companies, however, the creation of data that trains their own artificial intelligence is the gold that they need in order to sustain their company’s value. These companies create their own capability, their unique AI code, and their own platforms. We have witnessed singular companies create an entirely new world of platform economics: Think Airbnb, Facebook, and Alibaba—data is their asset.
Until recently, the modern miners of AI-driven insights had been left to create their own tools and workflows. As the community of data-economy prospectors grows, one might ask: Wouldn’t they be better off to focus on accelerating innovation for the primary capabilities of their specific AI-infused processes instead of building their own underlying infrastructure? Technology companies have the opportunity to create AI tools, leveraging technology platforms and business solutions. Even most data-centric companies could leverage modern AI tools rather than build them.
Typical of new innovation, there are always trailblazers who create their own unique solutions. However, most companies are not wholly ‘data-centric,’ and they focus on what they do best and delivering their brand promise. Most companies will not create AI from scratch. Instead, they will tailor AI off-the-shelf. This means using data as the glue, the trigger, and connector for intelligent processes. Think of this as similar to how companies today customize modern cloud ERP and business applications. In terms of AI, this means using AI-infused business solutions to create intelligent processes, or data being used to drive the intelligent enterprise.
What does it mean to make money by creating AI versus using AI? This is the parallel of prospecting for gold versus creating a new market—like blue jeans in the time of original prospectors, or a supply chain, living off the mining versus living off the miners.
Let’s illustrate this with examples of firms that use AI data to connect processes. Car companies can use data to predict when a car will need maintenance, ultimately using IoT-connected vehicles to inform maintenance of a needed repair. This can also to order or payment data—using data from unstructured invoices, forms, or emails to execute a service. Another example is the ability to act on stock-in-transit delays. There are a plethora of repetitive, mind-numbing tasks that could be offloaded from valuable knowledge workers, experts, and professionals in an organization.
Companies That Use AI to Connect Processes
Full-service hotel staff members answer many routine questions every day: Where is the hotel gym? When is it open? What’s the Wi-Fi password? What time is breakfast? Modern hotels today are testing and using ‘conversational bots’ to augment first-line service with typical answers. Rather than using staff members to answer, hotels use a bot which appears in an ‘ask-and-answer’ form such as a text service. The dialogue then connects back to staff with outcomes and more complex questions. The staff pick up then (connecting the service) and can perform more creative service to offer an excellent guest experience.
In all these cases, data is being used to create new processes that digitally enable and even transform an organization while AI helps to answer data questions.
Of course, these companies can “create AI” and become prospectors as well. However, most companies still spend too much time—and too many of their best resources—building up bespoke tools, machine-learning code, and custom data frameworks. This slows time to market and consumes resources that should be focused on driving innovation and sustainable differentiation.
There will be rare success stories, such as Uber, Amazon, and Netflix. Just like earlier gold rushes, including the original prospectors, a minority of players will make money from finding the gold (creating the AI and machine learning code, capability, and platform). But many more will become wealthy by living off the miners’ gold with food, supplies, services—or inventing new products and creating a market like blue jeans.
The real question is this: How will you capitalize AI?
Embrace AI, machine learning, advanced analytics, blockchain, and other emerging technologies to solve specific business problems before they are packaged for non-early adopters to procure later on. Get the insight you need by downloading the Economist Intelligence Unit study, “Making the Most of Machine Learning: 5 Lessons from Fast Learners,” sponsored by SAP. And explore SAP Leonardo, which is designed to allow businesses to take advantage of emerging technologies and merge them with their business data to innovate faster and transform their business more quickly – with less risk.
Marc Teerlink is global vice president of SAP Leonardo, new markets, and artificial intelligence at SAP.