These questions are based largely (though not completely) on runaway imaginations and jumping to conclusions about what the future will hold for these emerging technologies. Yes, they are important questions, and they need to be addressed, but, in many cases, it’s premature to design policy and legal frameworks around what might not be a problem, to begin with.
What’s lost in much of this is that the imagined scary applications of these emerging technologies are incredibly difficult to achieve. An extraordinary amount of software and hardware engineering work remains to be done, for example, before a fully autonomous automobile can be trusted to hit the streets in any but the most controlled, favorable environments. AI’s most common application, recognizing objects in a photograph, is still notoriously inaccurate and easily confused.
With that said, there are less-scary, less-challenging applications of emerging technologies that are making inroads in the enterprise. The promise of these applications is that they will make businesses more efficient, more agile, and more responsive to signals from the market, customers, vendors, and competitors.
Late last year in this blog, Abdul wrote about the concept of “digital transformation” and why it’s become such a big deal in the enterprise in recent years. In this post, I want to take the concept a bit further and discuss how emerging technology, specifically AI, is becoming an important part of many companies’ overall digital transformation strategy. It’s a growing trend that businesses ignore at their peril.
First, let’s review the concept of digital transformation and why it’s so important to the modern enterprise.
In the abstract, digital transformation is the process of reinventing existing business processes from scratch, leveraging digital technologies. From Abdul’s definition, digital transformation involves “organizational change through the use of digital technologies and business models to improve performance.” Under this definition, the three pillars of digital transformation are:
Understood this way, digital transformation is more than simply taking existing non-digital business processes (perhaps paper-based or manual) and creating electronic equivalents of them. It requires a hard look at how these processes, and the organizational structure around them, are constraining businesses from achieving short- and long-term goals, and determining which processes need to be automated—and which should be eliminated entirely.
That “hard look” needs to be done first—there’s no point in automating (or otherwise digitizing) a process that you shouldn’t be doing in the first place.
AI (and its various sub-genres, such as machine learning and deep learning) in many ways is still in its infancy, and its capabilities at present are quite limited. Training such systems, even to perform simple tasks, requires prodigious amounts of data, data storage capacity, and computing power.
Thus, as mentioned earlier, it is unrealistic to expect AI systems to remove the need for all human decision-making in business. However, well-trained AI systems, in general, are quite good at one thing: Recognizing patterns, including those that are beyond the ability of most human operators to identify, whether because of the amount of data to be processed or because of the subtlety of the pattern.
Because many business decisions can hinge upon the ability to recognize patterns, AI can be useful for businesses, even when applied to narrowly focused domains. For example:
In each of the preceding examples, the value that AI brings to the table isn’t that AI is “smarter” than humans—it’s not, now, and may never be considered “smarter.” AI’s value to a business rather lies in these benefits:
The drawback is that there is no “general purpose” AI system—if you want to apply AI to some task, you have to teach it that task, using huge quantities of data. Teaching an AI system to distinguish defective parts in photographs, for example, means “showing” it thousands of photos (each of which must be tagged as “has defects” or “defect-free” by humans) of the part in question. If the data doesn’t already exist, you need to produce it somehow.
Thus, the up-front costs of implementing an AI system can be quite high, but the benefits can be highly compelling.
Back in 1985, Harvard Business School professor Michael Porter identified three generic strategies that business use to attain competitive advantage: Cost leadership (minimizing the costs of providing goods and services), differentiation (distinguishing oneself positively from one’s competitors), and focus (concentrating on particular, narrow market segments). Not every successful business employs all of these strategies, but all successful businesses employ at least one.
These strategies are just as relevant now as they were over 30 years ago, but there’s a difference: Companies are now able to apply digital technologies to execute these strategies faster than ever before. In the modern business environment, speed is everything: Whether being first in the market with a product or being first to change directions when conditions warrant.
Old-style business organization, management, planning, and execution are no longer viable, and companies that insist on adhering to old approaches risk being left in the dust by their faster, more nimble competitors.
This is why digital transformation is so important. Applying digital technologies such as AI is the key to cost reduction, market differentiation, and market focus—that is, each of Porter’s generic strategies.
Digital transformation is not cheap and not easy. But in the long run, it’s cheaper and easier than watching your competitors outmaneuver you at every market turn.