AI for Your Business? Opportunities but Proceed Cautiously

September 2019
Artificial Intelligence technologies are still in their early stages of development. Despite the market hype, AI is complex and expensive to implement in many application areas, with difficult-to-quantify ROI. However, there are areas, with caution, where AI-enabled solutions are worth the effort for businesses or practices to pursue today. We’ll walk through some of the options for decision-makers to consider.
Artificial Intelligence (AI) is widely touted as a technology that will fundamentally change how we live, work, and relate to one another. There are AI applications everywhere you look. Autonomous self-driving vehicles, intelligent social media content curation, and smart home devices learn and adjust to our behavior.
With all this progress, should you act now to grab a competitive advantage with AI in your business or practice? The answer is – that it depends.
In this article, we’ll sort through the confusion about what AI is. We’ll provide you with an approach to sort the hype from opportunities. And we’ll explore how best to use artificial intelligence now while maintaining an appropriate measure of caution.
What Is Artificial Intelligence (AI)?
Artificial Intelligence means a lot of different things to different people. For some, it is a core enabler of the Fourth Industrial Revolution. For others, like Stephen Hawking, it could be a technology that dooms humanity.
However, most people haven’t come to an opinion on AI’s societal implications. Today, leaders are merely trying to figure out what AI means to their business or practice.
Artificial Intelligence is a sub-field of computer science. Its goal is to develop computer systems capable of performing tasks that usually require human intelligence. Its origins date back to a workshop held in 1956 at Dartmouth College. This is widely considered to be the founding event of artificial intelligence as a field in technology.
Three Different Goals for AI
The definition of artificial intelligence then shifts according to the goals of an AI system. These differences fall into three camps.
Strong AI. For some, the goal is to build systems that think exactly how people do. This effort to simulate genuine human reasoning is called “strong AI.” Researchers are actively working on this challenge. No one, however, has yet created a system that fully simulates human cognition.
Weak AI. For others, the goal is to get an AI system to work successfully regardless of how closely it truly reflects human thinking. This approach is referred to as “weak AI.” For example, when IBM’s Deep Blue beat world chess champion Garry Kasparov in 1997, the system did not play in ways similar to human thinking. The computer code used by the IBM engineering team didn’t include anything that resembled human thought processes.

The IBM Deep Blue Team in 1997: Chung-Jen Tan, Murray Campbell, Joe Hoane, Feng-Hsiung Hsu, and Jerry Brody.
Middle-Ground AI. There is a third camp that falls in between strong AI and weak AI. These researchers are developing systems that use human reasoning as a guide but don’t expect to emulate it perfectly. IBM’s Watson is an example of this approach. In the same way humans do, Watson can notice patterns in text that provide a little bit of evidence and then add all that evidence up to get an answer.
Google is also an excellent reference point. Google approaches AI by building artificial neural networks that mimic how human brains sort and process information. Their use of many layers of neural networks stacked on each other is known as a deep neural network and also exemplifies this “middle ground” AI approach.
The bottom line is that various forms of AI are being developed today, all smart but varied in how they approach developing intelligent solutions.
Hype Versus Viability
Fourth Industrial Revolution! This powerful concept was first introduced by Klaus Schwab in a 2015 article in Foreign Affairs, followed by a 2016 book titled with the same phrase. The name Klaus Schwab may not immediately make a mental connection for you. However, Klaus is a business professor at the University of Geneva who founded the World Economic Forum (WEF) in 1971.
Are you still stumped? The WEF’s flagship annual event is an invitation-only meeting of CEOs, movers, and shakers in Davos, Switzerland. Ah, now you’ve made the connection.
What is at the core of this envisioned revolution? According to Schwab, the “artificial intelligence already all around us, from supercomputers, drones, and virtual assistants…” will “fundamentally change the way we live, work and relate to one another.”
This is an exciting vision. However, we recall that similar words were used during the emergence of the commercial Internet in 1994 and with the arrival of personal computers in the late 1970s. But Professor Schwab’s enthusiastic AI vision has stimulated many global CEOs in his forum to integrate it into their strategic planning outlooks, even though most haven’t determined business cases for it yet.
Gartner’s Hype Cycle
Market hype is so typical today with emerging new technologies that Gartner, a leading research and advisory company, created a “Hype Cycle” framework (Exhibit 1) to help business leaders discern hype from what is commercially viable. Of the 37 technologies that Gartner follows in the artificial intelligence technology sector, only two have reached the plateau of productivity in its 2019 market assessment: speech recognition and GPU accelerators. Gartner notes, “[2019’s] Hype Cycle features many new technologies, but few have a value or purpose that is fully understood.”

Exhibit 1: Gartner’s Hype Cycle framework, providing a graphic representation of the maturity and adoption of technologies and applications and how they are potentially relevant to solving real business problems and exploiting new opportunities.
Forrester, another leading research and advisory company, states that the AI journey of large enterprises will be a “rollercoaster.” In their “Predictions 2019: Artificial Intelligence” report, they emphasize that pursuing AI will be risky without immediate rewards. Of core importance, they emphasize, is for leaders to plan meticulously, fail fast, and continue to learn more about implementing AI from that process.
The Early Internet Era Provides Lessons for Today’s Leaders & Executives
In the words of the great philosopher Yogi Berra, “its déjà vu all over again.” When corporate America began to sort through the implications of the Internet in the mid-1990s, CEOs would pick up cocktail conversations as excitement about the Web started to build. They’d return to their staffs with a mandate to get Internet initiatives underway for their companies and organizations. However, there typically was no guidance from them about why and for what specific purposes the Web should be used.
The impetus for those CEOs was grounded in fear of being left behind by their competitors. These CEOs could sense but not put their finger on the specifics of the Internet’s transformational nature. Their intuition told them that the Web would foster significant opportunities to change customer behavior and business-to-business relationships, even though they did not understand precisely how. As a result, corporations spent money on Internet technologies with abandon, but a tangible return on investments in those early years was elusive.

Lou Gerstner, chairman of the board and chief executive officer of IBM from April 1993 until 2002
IBM’s Lou Gerstner confronted this widely adopted approach with his keynote presentation at the COMDEX computer expo in Las Vegas in the fall of 1995. He summed it up quite succinctly: When executives were considering investing in Internet initiatives for their companies, “It’s about business, not technology.”
When adopting AI today, just as it was for eBusiness back then, you need to lead your efforts with viable business objectives that justify your investments. Only the largest enterprises have the resources to throw at AI and learn through iterative failing efforts. Most businesses and practices, however, can’t afford to operate on a “wing and a prayer.”
Realistic Expectations for Potential AI Projects
You need to set realistic expectations for AI. According to Gartner’s 2019 CIO Agenda survey, only 14% of large enterprises have deployed artificial intelligence initiatives, up from 4% the previous year. You are an anomaly if you are a small to midsize business or practice using AI other than on an experimental basis.
This is not to say that AI can’t have a transformational impact on a business of any size – but leading analysts are predicting that the commercial viability of the next wave of AI technologies won’t be for another 2 to 5 years (aside from speech recognition and GPU accelerators, as we previously discussed). Implementing AI that drives business value requires talent with highly specialized skills and involves significant expenses. That talent is in short supply, and many others with AI skill sets fall short but are still eager to take advantage of growing market demand.
Like all new technologies, AI will become less expensive, have better-defined business uses, and require less computer science knowledge to implement over time. Given the current hype and excitement around AI technology, it is understandable why many small to midsize business and practice leaders feel they should be finding ways to use AI immediately. But for most, AI won’t help to achieve any clear business objectives, and it is a technology that requires time and skills, making it a challenge to implement effectively.
Four Core Considerations for Decision-Makers
Summing up our core considerations for enterprises as well as small to midsize businesses and practices concerning AI investments:
- Save for the exceptions outlined below, don’t consider using AI technologies until they are commercially viable. At this point, the costs will have fallen considerably, and enabling technologies’ sophistication will have evolved significantly.
- If you are compelled to implement an AI solution now, see it through by committing the time and resources required to implement it with quality. If you become overwhelmed with the time and costs of an AI project and decide to launch it prematurely, it will negatively impact your brand with your customers and clients. As Warren Buffett has noted, “It takes 20 years to build a reputation and 5 minutes to destroy it.”
- You should stay on top of developments in AI technology, and start building AI business cases in the context of your business goals now, so you are ready to move ahead at the appropriate time.
- There will be a few business leaders who see a path to leverage AI technology to improve business functions and profitability now. If you are this kind of visionary, and your planning is detailed and sound, press ahead.
Opportunities to Experience AI “Off-the-Shelf”
Even if you don’t implement AI directly today, you have plenty of opportunities to experience it. Many current “off-the-shelf” technologies and applications use artificial intelligence in their programming. For example:
- Microsoft Word is developing an Ideas feature for its online version to suggest grammar changes, among other assists. In addition to catching basic errors, it can recommend rewriting phrases to improve conciseness, clarity, and inclusiveness.
- Salesforce’s Einstein features use AI to help determine which channels, messages, and content resonate with your customers by automatically analyzing their interactions with a business or practice across all customer touchpoints.
- Constant Contact uses AI to determine the best time to send a newsletter so that it will be opened based on reader patterns in your database as well as in similar industries.
- Facebook and Google Ads use AI to help you construct an ad consistent with the messaging on your landing page or website and contains the keywords that will help search engines match your ads to
Exercise Caution with Chatbot Hype
Simple chatbots have been around for years and are computer programs designed to simulate conversation with the humans using them, particularly over the Internet. They are built using pre-written keywords, regular expressions, and forms of string analysis. They are relatively dumb save for a developer’s extensive work to create them.
Smart chatbots rely on artificial intelligence when they communicate with human users. Instead of pre-prepared answers, smart chatbots can learn and improve beyond their programming by discovering patterns in data. They can later apply those learned patterns to subsequent conversations and adjust to slightly different questions. This ability gives them the “intelligence” to perform tasks, solve problems, and manage information without human intervention.
The upside to smart chatbots is that they can lower the barriers to continuous user engagement. They can improve customer service by being immediately available and automating repetitive inquiries. They can personalize user experience and reduce the workload on your customer service staff.
In recent years, vendors have made an avalanche of technology available to help the smallest businesses develop their chatbots. The problem is that creating a chatbot successfully that works robustly, doesn’t irritate your customers, and serves a business objective well involves much complexity and work to implement. While the cost of the technology to operate a chatbot is relatively small, the expense and complexity of launching a quality chatbot can be significant.
We continually run into cases of leaders launching chatbot projects today just because the technology is there. Don’t fall into the same trap. Only move forward if you have a business case that makes sense, you have access to the expertise that can implement a chatbot for you without detracting from your brand, and you are prepared to invest the time and resources required to achieve quality results.
For most small to midsize businesses and practices, the best way to leverage artificial intelligence today is in off-the-shelf applications and technologies where it is embedded. For large companies, investing part of your research & development budget into AI may be appropriate now if you can link such an effort to future business models.
You shouldn’t take on AI projects because you feel pressured to do so – you should only move forward with a business case that will justify the required time and significant investment. In the interim, you should stay on top of developments in AI technology and start building AI business cases in the context of your business goals now. This will be key to your readiness to move ahead at the appropriate time when AI is less expensive, has better-defined business uses, and requires less computer science knowledge to implement.
About the Author
Ted Bream is the Managing Partner of Core Agenda. He is passionate about helping leadership teams figure out how to leverage the growing assortment of digital, technological, and creative tools available to grow and sustain their market presence. His priority is getting those initiatives effectively synchronized within the context of each client’s business goals and operational realities.
Among its work, Core Agenda is currently developing early-stage AI initiatives with its clients.