It’s one of the paradoxes of emerging technologies that hype and reality rarely match. Artificial intelligence is a case in point: the excitement about its likely impact on the business world continues to build while in the real world, this very powerful technology is having only a marginal effect on day-to-day operations. What can business leaders and technologists do to close the gap? I think there are 4 steps we can take:
1. Learn the building blocks of AI Algorithms
All the cool artificial intelligence applications featured in the press and promoted by vendors appear magical and mystical, but the magic is illusory. Yes, AI is very powerful and can achieve interesting results, but it’s not Harry Potter -style wizardry. One of the biggest challenges in converting AI visions into reality is bridging the knowledge gap between business people and data scientists. Business people understand the commercial landscape but not how AI can be applied to solve existing problems or to create new opportunities.
Meanwhile data scientists have a profound understanding of AI’s capabilities and limitations but are generally less knowledgeable about the imperatives of running a competitive business. Currently the gap between these two disciplines is so wide it’s difficult to have a productive conversation, so it’s crucial that both sides take responsibility for improving their knowledge about business/AI by talking to each other and using each other’s expertise to solve problems together.
This doesn’t mean business leaders should learn how to write Python code – it requires they gain an understanding of how different algorithms support what is possible as well as what the pitfalls and the trade-offs are. Developing this technical understanding enables business people to envision and innovate solutions to their business problems and also empowers them to challenge their technical teams in a constructive way.
2. Understand the Critical Role of Data
One of the benefits of business leaders improving their technical understanding is that it helps provide them with much needed insight into the role of data. Just as “power is nothing without control” (Pirelli tyre ad), AI algorithms are nothing without data. Or as machine learning guru Pedro Domingos puts it: ”a dumb algorithm with lots and lots of data beats a clever one with modest amounts of it.” After all, the quality and quantity of the data helps define how well AI can predict things. This explains why 80 % of time spent on AI projects is spent working with data: getting access, cleaning, preparing, pre-processing, normalising: the grind that precedes the more ‘fun’ stuff.
Data becomes a strategic issue when the right data isn’t available at the right time. Creating and collecting the required data set in this scenario is costly. The strategic importance of data means it’s essential for business leaders to understand the role of data, what kind of data they already have and what they still need, so that work on building the right data set gets underway immediately.
3. Define Your Ambition Level
The most obvious real-world AI applications include automating processes and innovating efficiencies. Other typical applications include providing better value to customers through value-added features, such as intelligent products that learn users’ preferences and make personalised recommendations. These applications are a good place to start. They are relatively straightforward to execute, and ROI is easy to quantify.
While it’s important to start small, business leaders should also explore more far-reaching opportunities, such as using AI to help identify new revenue sources, expand to new industries and holistically redefine what business the company is in.
So, what would it look like in practice?
- New revenue source: Monetising harbour data about ship and cargo movements to provide investors with insights about the global economy.
- Expanding into new industries: E-invoicing providers turning into financing businesses based on performance and risk.
- Redefining business: Using AI to shift the business model from producing goods/services to selling knowledge about how those goods/services are produced.
When it comes to deciding how to implement AI, business leaders need to define their level of ambition: is the goal to improve existing business operations or to build a completely new business?
4. Be Aware of AI’s Cultural Impact
Digital transformation has been an important catalyst for shifting companies away from a hierarchical, siloed, top-down organisation towards a more open and creative bottom-up culture.
Conventional management strategy is about planning, certainty, hierarchies, functional silos, incremental innovation and execution. Innovation strategy is about experimentation, collaboration, empowerment, creativity, freedom and failing fast. Now data and AI are driving a third cultural paradigm in which automation and data-driven facts trump opinions and where probabilities are used to address uncertainties. This is an era of human-machine collaboration which will require a rethink of traditional operating models, role definitions, individual success measures and career progress.
For businesses with a high ambition level, understanding AI’s cultural impact is key because it’s not possible to change gears operationally without also changing a company’s cultural mindset around its core values and purpose, as well as how it goes about identifying and solving problems for its customers.