Two businessmen arguing over who owns an AI project
PHOTO: Shutterstock

Digital transformation is a fact of every business today. While the transformation of business has been well documented, it is less clear how technology, and what technologies are going to drive this process. Recently, research from Grant Thorton, the world's fifth largest professional services network of independent accounting and consulting member firms based in the London, UK, has quantified how much digital transformation will cost, how it will impact financially on enterprises and what the ultimate shape of digital enterprises will be.

The research, which was published at the end of February, surveyed 304 CFOs and other senior financial leaders, from companies with revenues between $100 million and over $20 billion. Overall, more than three quarters of the executives surveyed agreed that digital transformation is critical, 23 percent in the short-term and 56% in the long-term. The data also showed digital transformation investment is on the rise throughout the enterprise, increasingly focusing on strategic opportunities. While organizations are moving forward gradually when it comes to adopting new technologies, they do so at a steady pace to both enable and disrupt current and future business models.

AI Top of Mind For CFOs

While the report showed that in the future, customer experience tops the ranking in terms of investment priorities, it also showed that competitive differentiation was one of the top four goals. To ensure that differentiation, artificial intelligence (AI) was a top-of-mind topic for CFOs.  The research showed that 21 percent of CFOs noted their organizations plan to adopt artificial intelligence technology in the next five years.

It also found that future investments in automation will focus on better strategy, analysis and predictions for growth, which explains why executives identified advanced analytics and artificial intelligence, as top projected investments in new technologies over the coming years.  Even though the investment plan is there, they differ in how the technology will affect their staffs.

  • 52 percent plan to train existing staff
  • 20 percent will recruit new staff
  • 17 percent prefer to outsource with a third party

Related Article: Why the Benefits of Artificial Intelligence Outweigh the Risks

Ensure the Right People Are Running Your AI Projects

Hyoun Park, CEO and Principal Analyst at Amalgam Insights points out that AI is being treated first and foremost as a progression of business process management and automation, so that the line of business staff previously responsible for enterprise applications and application integration are being placed in charge of AI. However, the challenge here is that these personnel often lack knowledge on how some aspects of AI work. “The guided machine learning, data science, and deep machine learning associated with true AI requires a knowledge of statistics, algorithms, and coding to understand the potential outcomes and answers that AI can provide. Companies seeking to understand how AI works should be placing their AI efforts under the watch of a CTO or a Chief Analytics Officer who understands both the data and related calculations within AI,” he said.

However, there are also text and graphic AI efforts that are more focused on creating chatbots, natural language generation, natural language processing, and graphic matching in pictures and videos. For these efforts, the focus needs to be on providing outcomes that are more human and less biased in nature. This will require a guided human approach that should be led by skilled sales, marketing and communications executives. 

In these efforts, tone, context, metadata, and feeling are vitally important. “Taking a purely technical or process driven effort to text and graphic analytics and AI will result in taking algorithms at face value when they shouldn't be. All AI is still relatively naive and stupid at this point. It doesn't mean the AI can't be effective or work at massive scale, but AI lacks human levels of judgment and general context which lead to the need for guidance,” he added.

As a result, AI and how it is used in the enterprise should not be left to C-Suite alone, according to Chris Butler, director of AI at Palo Alto, Calif.-based Philosophie, which works with brands, agencies, and startups to create usable, meaningful digital experiences. He says that while the C-Suite should be driving strategic direction, hiring, and other executive functions, the implementation details of which technologies to use should be a discussion between product, design, data scientists, engineering, and operations. “AI isn't just magic fairy dust that you sprinkle on projects because someone upstairs tells you to add it. It can take a significant investment in data collection and cleaning, model specification and training and operational monitoring,” he said.

Related Article: 8 Examples of Artificial Intelligence (AI) in the Workplace

The Need For AI Ownership and Leadership

However, at some point in the process, one person needs to take charge of the process. For Cristian Rennella, CTO and co-founder of elMejorTrato.com that person should be the CTO. He argues that there comes a time when the success of AI is key to the company. Not only will it be used in sales, but also in marketing and in human resources. That is why, the CTO must take this responsibility, in the same way that he or she manages the systems of the company. For any of these departments to grow, AI must be incorporated into the enterprise IT infrastructure

Dylan Max is  growth marketer lead at San Francisco-based FogLogic. He said that much of the AI work is functional, meaning that it needs to be owned by the business function, facilitated perhaps by IT depending on the interfaces. For the most part, AI will solve or help manage business problems, so the functional owner needs to be responsible for the AI just as much as they are for people making decisions. “Infrastructure may be shared, and there may be technology standards to adhere to, but the logic, the reason and the inferences still belong to the function owner,” he said. “AI and IoT are implementations, not ideas, so they need to be as close to the ground as possible so that means the manufacturing team, customer service, logistics. Whatever function(s) will benefit, need to own the implementation."

Pull all this together and it suggests that companies looking to make a strategic commitment to AI across their business should create a position focused on AI implementations, like the way many companies vest responsibility for information security in a Chief Information Security Officer — and the position should report to the CIO. An alternative approach that could be effective in companies that have designated a Chief Digital Officer is to have the Head of AI Strategy report to that role.

Invest in Customer Research and Prototyping

Butler says that organizations need to do their due diligence if they want to be successful with AI implementations. “You should invest in prototyping and customer research before you build whole solutions. We have found at that it can save months of work if we understand the potential impact of an AI solution before it is fully built," said Butler.