The transformative potential of artificial intelligence is a topic of discussion in a growing number of boardrooms, classrooms, parliaments and public spaces around the world. It’s little wonder, given that a recent study by PwC found that AI could generate 14% growth in global GDP by the end of the decade, worth a combined US$15.7-trillion. How is that possible? Because of the productivity gains AI can underpin, and the demand for new services and products it can drive.
The same study expects productivity to grow by as much as 40% thanks to AI implementation, and not only in the private sector. Governments are realising its outsized potential for the public sector too. At its core, AI is about imbuing computers with human-like powers of cognition, pattern recognition, and the ability to learn new skills; and it’s getting better all the time.
AI can increasingly learn human things, but unlike humans, its knowledge can be immediately distributed across multiple instances, and something learned by one implementation can be rolled out to other implementations in almost real time. In that sense, it’s cumulative and thus, its advances are exponentially accretive and rapid.
Executives see it as a way to boost productivity, serve customers better, and anticipate and capitalise on new business opportunities. A global survey on AI by McKinsey conducted in 2020 found that nearly half of business leaders who responded had adopted AI in at least one area of their business, whether in human resources, corporate finance, manufacturing, supply chain management or other verticals.
Eventually, AI will transform every industry, and every business within them. In 2021, McKinsey says it saw $165-billion in investment into AI across a plethora of industries, and according to the IDC, it expects expenditure on AI-related hardware and software to reach $97.9-billion this year alone.
But if governments and businesses are to capitalise on this enormous opportunity, they need to have clearly thought-out processes and structures, articulated by comprehensive AI strategies.
An appropriate AI strategy helps a business focus on its core objectives and prioritise how and where AI can assist it in realising its goals. A strategy also articulates how a business measures value, and how to identify where it is in its AI maturity cycle. As an example, as per Gartner’s AI Maturity Cycle, most organisations today are at the awareness or active stages: that is, they’re interested in AI and may be experimenting with it. At the far end of the spectrum are organisations where AI is ingrained in the business DNA and transformational.
A Forbes AI survey from 2020 found that among respondents AI initiatives increased to 64.8% in 2020 from a mere 39.7% in 2018, with a total of 98.8% of firms investing in it in one way or another. A full 73.4% of firms cited big data adoption as an ongoing challenge, while 90.9% of respondents cited people and process challenges as the biggest barriers to adoption and success.
Perhaps most surprising, according to Forbes, only 14.6% of organisations reported that they had deployed AI capabilities into widespread production. That lack of successful deployment and the challenges cited are prime examples of AI obstacles that can be overcome with an appropriate AI strategy.
In 2019, the Harvard Business Review conducted a survey of its own focused on “Building the AI-Powered Organisation” and found that of the thousands of companies and executives it spoke to, only 8% engaged in “core practices that support widespread adoption.” Many companies, it appears, struggle to see AI as a potent tool and a means to create meaningful impact, rather than the impact itself. Again, a suitable strategy can alleviate this and help a company drive adoption and make the most of what AI can do while avoiding the pitfall of expecting the impossible from it.
Like any new process, technology, or tool that’s transformative, to get the most out of AI you need to focus on objectives and not on the novelty of AI. An AI strategy ensures the focus remains on outcomes and aligns them to have a decisive impact. Furthermore, implementing AI doesn’t just mean hiring data scientists — talented practitioners in the AI space want to work for companies that understand the technology and have a strategy for how best to leverage it.
As Murat Durmus, the CEO of AISOMA, succinctly puts it: “A company nowadays without an Al strategy is like a sailboat without a sail.”
A framework outlined by Jacob Bergdahl published in “Towards Data Science” (and by other AI experts) shows the role of AI today tends to fall into two broad categories: automation and augmentation. Automation is when AI takes over an activity completely. Augmentation is when AI assists or aids in an activity, empowering the person carrying it out in the process. Many AI tools sit somewhere between the two. Between these two is a scale of four strategies:
- The efficiency strategy: activities are optimised by means of automation
- The effectiveness strategy: activities are made more seamless
- The expert strategy: AI empowers and complements decision-making
- The innovation strategy: AI enables enhanced creativity
An AI strategy can help a business identify which of the above strategies are best suited to the business problems it’s hoping to solve. Moreover, it can help identify those very problems. Activities with low data complexity are often ripe for automation. High data-complexity ones, meanwhile, tend to be better suited to augmentation. But, of course, there’s some overlap depending on the specific use case.
Moreover, by outlining a clear AI strategy, you can get a glimpse of what a mature strategy and cultural transformation could look like down the road and get a sense of what it’ll take to get there.
- The author, Dr Mark Nasila, is chief analytics officer in FNB’s chief risk office
- Read more articles by Nasila on TechCentral
- This promoted content was paid for by the party concerned