At several recent conferences, including RoboBusiness 2017 in Silicon Valley, sessions tried to demystify the hype around artificial intelligence. In many conversations, speakers and attendees asked the same question: “How can AI applications help my business, today and quickly?”
This may seem like a strange question, given all the talk about AI, but it is not. There are two reasons for that.
First, few executives properly understand machine learning and related technologies, what problems they need to solve, and what is possible. Most people in their leadership teams have formed opinions based on anecdotal stories.
Second, both startups and established vendors are launching solution after solution. Every rollout is called an “industry first” or the best thing since sliced bread. Basically, there are too many hammers looking for nails.
When you these two explanations together, it is not surprising that many executives are confused and paralyzed about AI applications.
“AI” is an umbrella term, and there are a lot of different flavors. Hence, there is no one-size-fits-all approach. Depending on the business problem, even within a single company’s operations of a single company, you may end up with different AI applications, from bots for customer service and virtual financial analysts to predictive maintenance and route planning.
So, how can you profit from AI, now and fast?[note style=”success” show_icon=”false”]
- The global market for AI applications will have a compound annual growth rate of 50.51% from 2017 to 2021, predicts Orbis Research. The market will reach $35 billion by 2025, says Grand View Research Inc.
- Business leaders should know whether they’re using AI to expand an existing capability or add a new one, and the technology presents an opportunity to alter customer relationships.
- Another area in which bots can help is in augmenting the workforce and improving productivity.
Whether it is a bot reading e-mails or a more advanced virtual agent that can understand voice and emotions, AI is going to be your next worker. Instead of thinking, “Which parts of a process can a bot do?” ask yourself: “What capabilities should I invest in and develop that can scale as required?”
You will still use AI, but the way your frame the question will make a huge difference on success.
That’s the approach that SEB, one of the largest banks in Sweden, adopted. SEB knew that by implementing a “cognitive agent,” it would benefit from greater efficiency. However, the bank also wanted a capability to deal with customers that it could scale as required, quickly and cost-effectively.
You are going to have to “teach” AI applications a lot about your processes, data, and business model. It is no different than when you bring a new human employee on board, so why not approach AI the same way?
How many customers you work with? One, 1,000, 1 million, or perhaps 100 million and more? What’s the one thing that could redefine your relationship with them?
Is it the ability to offer a timely discount coupon after analyzing data and predicting behavior? Or is it providing customer support that people talk about in cafes and cocktail parties? Or, is it something else?
If you cannot answer this but set up a chat bot out to “automate” and “streamline” the process, don’t be surprised if your bot ends up being a dud. Just on Facebook’s Messenger platform, 70% of the bots have failed.
Bots are the fastest-growing AI application — for now. But you do not have to start with bots to lead the customer.
KFC has partnered with Alibaba in China to use facial recognition software to process payments. Today, you can “smile to pay.” Tomorrow, you may be able to get a selfie “for free or fee” — another creative way to lead the customer.
Once this new payment method scales up, KFC may end up with hundreds of millions of pictures in its database. The next-level possibilities to identify new revenue streams and build new forms of partnerships with customers are enormous.
Before you rush toward a specific vendor or a popular AI application, make sure you know how to lead your customers with AI.
Every business is looking for new talent, new skills. Companies are rushing to try out bots to please customers or automate routine work. But, one obvious area where these AI applications can have a dramatic impact is in workforce automation.
Why not create a bot assistant for every employee in your business?
When Australia’s next school year starts in February 2018, Deakin University will provide a bot to all of its 56,000 students. The bot will be a coach as well as a companion.
Providing AI assistants to every employee offers benefits to businesses of any size. People will learn to coexist with robots (in the broad sense). Staffers can connect with roles and goals a lot better — “cool” is always exciting. Of course, you will also see productivity improvements, thanks to bots tracking and guiding everybody.[note style=”success” show_icon=”true”]
More on AI Applications:
- Can AI Improve Early Breast Cancer Detection?
- Military AI Provides a Preview of Commercial Capabilities
- AI System Could Help Robots See Around Corners
- AI Readiness Key to Business Advancement, Not Just Tech Status
- Toyota Robotics, AI Investments Are Selective, Strategic, Says RoboBusiness Panelist
- CROs Need Robotics, AI Roadmap to Prepare Others for the Future
- Introducing Industrial IoT to Manufacturing — Challenges and Benefits
- Healthcare AI to Be First Point of Contact for Patients, Policymakers
Getting started with AI applications
None of these ideas is too radical. You can begin with any that makes sense to you, or you may have an idea or two of your own.
It does not matter which idea you begin with. The key to profiting from AI is, first, having strategic clarity — knowing why you are automating, what is the business problem, and what your organization wants to get out of it. You’ll need lots of patience, since there is no overnight success, lots of mistakes will be made, and there will be times when you will wonder if you made the right decision.
Technology is the easy part. Turning technology into treasure requires work.