“Everything invented in the past 150 years will be reinvented using AI within the next 15 years.”—Randy Dean, Launchpad.AI
Where would you say you are in the process of artificial intelligence (AI) adoption in your organization? While pioneers at the world’s most innovative and successful organizations are doubling down on AI and data analytics, many other enterprise organizations remain clearly behind the curve. Global AI spending is expected to reach $35.8 Billion in 2019, according to International Data Corporation, and AI is expected to add $15.7 trillion in value to the global economy in 2030. Is your organization prepared to participate?
We have talked before on this blog on how there is a growing gap between the AI “haves” and “have nots.” The evidence continues to mount that the organizations that are actively building AI capabilities are beginning to outpace those that don’t in significant ways. Yet, PwC research indicates that just 20% of executives say their companies are deploying AI across the business in 2019. What’s holding the other 80% back?
Here are seven ways that organizations can get started with AI:
1. Set clear goals.
“Let’s just get started” and “We know we need to do something with AI” are comments that should set off warning bells. Although some organizations treat AI adoption as an IT project, it’s more like a science experiment. There’s no time for “playing around” with AI if your competitors are further down the road. Your organization should have a clear goal in mind and define what success looks like at the outset. As in any experiment, a documented, measurable hypothesis will help to clarify your thinking about what outcomes count as a win.
2. Place small bets.
Regular readers of this blog know that we at SU are huge proponents of moonshot thinking. It’s important to think exponentially, but it’s also important not to interfere with your organization’s current value creation. For example, you may choose to test an AI-powered chatbot on a small segment of your customers before rolling it out on a larger scale. In the same way, you may choose an internal problem to solve that won’t have revenue or customer impact if things go sideways—which should be accepted as a potential outcome, especially early on.
3. Build and leverage core expertise.
A good way to make your team more comfortable with AI adoption is to solicit the advice of your organization’s domain experts in identifying and understanding current challenges. An incomplete understanding of the problem you’re trying to solve will lead to an incomplete solution. Getting input and ideas from your current domain experts can help you understand pain points and improvement opportunities more deeply from the start.
4. Pursue quick wins.
One of the best ways to gear up for solving massive problems is to first solve a small one. Look for a problem you can solve for a quick win. If you’re able to solve an existing pain point by eliminating a slow, expensive, or error-prone process, your credibility as an AI practitioner will soar. This is an opportunity for your organization’s domain experts to collaborate with AI and data science specialists to identify obstacles that are good candidates for streamlining or reinvention.
5. Clean your data.
We’ve talked about how to combine big data and AI for powerful results, but having clean data available may be even more important for organizations taking their first steps with AI. Many AI projects have fallen flat because suitable data sources were not identified at the outset. You may be excited about tackling a large data set that’s rich with potential insights, but if that data is incomplete, full of duplicates, or incorrectly formatted, you’ve got some data prep to do—which can get in the way of your quick win.
6. Use prepackaged solutions.
If yours is an enterprise that’s further back on the digital adoption curve, there is some good news. Today there are more data sets and off-the-shelf solutions available than ever before. A quick look at Amazon Web Services shows a variety of prepackaged solutions for AI and machine learning, but also analytics, blockchain, Internet of Things, and mobile applications. You can search for solutions by industry or by use cases. A huge advantage of using Amazon or comparable web services from Google Cloud or Microsoft Azure is that you won’t have to set up your own infrastructure and you’ll often have the flexibility to choose a DIY approach or purchase fully managed services.
7. Choose your champions.
Obviously, a huge part of making AI work for your organization is to have strong executive support. The buzz around AI and machine learning is certainly strong enough to reach the executive suite. But there is often a gap between appreciating AI’s massive potential and approving the budget and resources for a project. A shorter experiment with a realistic budget and clear timeline is more likely to be approved than a larger, more costly transformation effort. Strong support from the C-suite and boardrooms can mean the difference between success and failure for your project.
Take the next step
At SU, we often say that exponential enterprises are built on people and culture. Getting the right people on board at the outset is critical. Learning about emerging technology and leveraging its capabilities is the easy part. The human and cultural aspects are more challenging and unpredictable. For a deeper dive, learn more about how enterprise organizations can build powerful AI capabilities.
Companies around the world are coming to understand that AI is no longer an option, but a requirement for success. Successful AI and machine learning outcomes require each organization to build capabilities that enable them to move quickly from concept and ideation to production and profits. The knowledge, skills, and methodologies required to meet these challenges cannot be entirely outsourced.
Building your AI capabilities now will help position your organization to move faster in the future and avoid wasted time and resources. Even better, a strong AI capability in your organization enables your team to understand and conquer challenges that appear unsolvable to less-savvy competitors.
That’s because one thing that has not changed in the age of exponentials is that early adopters capture the lion’s share of opportunities and business value. And just as technological advancements are accelerating, so is the need to move quickly to outrun disruption and achieve competitive advantage.
Pro tip: Singularity’s new partnership with INSEAD is kicking off with a Future of AI program in January, 2020 dedicated to helping you face fears about AI head-on and better understand its use cases.