How to Get Started Developing a Corporate AI Strategy
Jul 30, 2019
By Jamiel Sheikh
Artificial intelligence (AI) is a not a cure-all solution to every corporate problem. It’s a rapidly emerging technology capable of making a significant difference to your bottom line, but AI still has a long way to go before it can provide the solutions that many organizations expect it to deliver. For very specific tasks, artificial intelligence performs better than humans. The question then becomes how to align strategy to capture value that AI can potentially deliver. At the most basic level, cost-reducing automation should come to mind. But are there revenue-generating opportunities? That depends.
Is Your Business Ready for AI?
An important first step is to figure out where it makes sense to integrate AI into your corporation. Although software companies may describe AI as being able to handle any number of tasks, the reality is that it takes time to integrate AI effectively into an existing business. Choosing one area and focusing on AI integration is a good first step toward developing a coherent strategy. AI is very good at handling repetitive or boring tasks and looking for a way to automate these tasks in your company is a good way to begin.
Arguably, the best use of AI is in the analysis and interpretation of Big Data. AI programs are brilliant at finding patterns and correlations in diverse data stacks, and it’s here that many businesses can see a marked improvement by adopting an AI system. Of course, for this to be effective, AI requires good data with which to work. That’s why it’s important for corporations to ensure that data collection is being handled efficiently. In creating a corporate AI strategy, it pays to optimize existing computer systems before laying out capital to integrate AI.
Focus on Low-Hanging Fruit of AI
AI is still coming of age, despite its recent acceleration in development and prominence in the media, and the noise can be overwhelming. I prefer to simplify things and look at low-hanging fruit when a market gets noisy. One area where it makes perfect sense to integrate AI today is in customer service. AI is still years away from passing the Turing test, where interaction with machines cannot be differentiated from interaction with a human, but for simple inquiries, customer service algorithms continue to improve and provide deeply contextual assistance to a customer. This actually creates an interesting situation—if an AI customer support system is answering questions via a chat bot, or some other means, should the company tell the consumer that they’re interacting with a computer? On the one hand, there may be the fear that the customer reacts negatively as they perceive that their inquiries are not being taken seriously. Despite this possibility, some businesses have actually experienced an increase in customer satisfaction after they’ve integrated AI into their customer support systems. This happens as a result of a customer’s tendency to be more honest with a computer system. When interacting with a computer, some customers may be more likely to ask inane questions because they’re not worried about a customer service representative judging them. Likewise, customers may also ask more questions because they’re not worried about taking up someone’s time.
In the end, the customer gets all of their questions answered, even the silly ones, and they don’t feel rushed. Although this is certainly not always the case, when well executed, an AI customer service system can be a great addition to any corporation and, perhaps most importantly, positions the organization to evolve as AI evolves.
As a tool, AI is better suited to improve upon the existing strengths of a company rather than being used as a method of drastically cutting operational costs and eliminating headcount. While it’s true that AI can do certain things that may make some employees redundant, its real power lies in the co-creative process. That is, pairing AI with people who understand how to use it, augmenting algorithms with the minds of people and augmenting the minds of people with algorithms. Some businesses may choose to improve marketing by using AI to detect consumer preferences. Others may look to boost product sales by having AI search for patterns in previous customers to determine who is most likely to buy—called recommender systems. Companies can even apply AI to their wider business strategy to see if it can detect any profitable positions that may have been overlooked—a large undertaking but potentially using algorithms to squeeze out extra return on investment.
In all three of these cases, of which these are just a few examples, the real capabilities of AI are far-reaching, the AI system is used to enhance the abilities of employees, not make them obsolete. While the AI of tomorrow may be capable of reasoning and intuiting in a humanlike fashion, the AI of today is not, and there are many, many things that only a human can do. Narrow AI, which encompasses all of AI today, are algorithms that perform specific tasks. General AI, the ability for algorithms to perform multiple human functions simultaneously, is still very far away from being a reality.
While the returns on a well-implemented AI system may be significant, there are expenditures that must be taken into consideration. The primary cost is recruiting and keeping on payroll talented engineers and programmers to build efficient AI systems, pulling together an effective technical strategy that maps well to the business objectives and managing performance of the initiative. Software engineers with AI experience can command relatively high salaries and all of this needs be included in your net present value and internal rate of return calculations.
Determine a budget and then figure out whether it makes more sense to hire away established engineers, or seek those with less experience and foster their talents in-house. Whichever option you choose, it makes sense to think about costs up front so that your company is not surprised by any expenses down the road. Artificial intelligence has the potential to pay for itself many times over; however, it can take time to realize those profits while in the meantime burning lots of cash.
Developing a corporate strategy for AI requires significant due diligence, potentially requiring multiple parts of an organization to come together. Start with some of the low-hanging fruit in AI with relatively immediate returns, like customer service. Doing so gets you on the AI train and allows you to start to accumulate actionable data about your customer. Second, find ways to augment AI into existing business processes, supplementing what humans do, allowing them to focus on more meaningful work. If you are a very large enterprise, set up an AI innovation lab and staff it with bright minds to help facilitate the process of finding where to augment first and how. The AI innovation team should be able to assess the ripest business processes for AI. Finally, be realistic about what investment AI may require and assess if the risks and upside are justified. It’s not cheap, but if done right, this can strengthen or establish sustainable competitive advantages.
About the Author(s)
Jamiel Sheikh is CEO of Chainhaus, an advisory, software development, application studio and education company focused on blockchain, artificial intelligence and machine learning. Jamiel has more than 15 years of experience in technology, capital markets, real estate and management working for organizations like Lehman Brothers, JPMorgan, Bank of America, Sun Microsystems, SONY and Citigroup. Jamiel is an adjunct professor at Columbia Business School, NYU and CUNY, teaching graduate-level blockchain, AI and data science subjects. He runs one of the largest blockchain, AI and data science Meetups in NYC.