Making Artificial Intelligence (AI) real in an organisation

Buoyed by AI conversations in the market place and number of strategic AI programs that I am personally involved in, I am convinced that AI is a mandatory transformative enabler for businesses to be successful in a world reshaped by emerging technologies, changing consumer behaviours and need for speed.

PWC’s 22nd CEO Survey had a reassuring observation at Exhibit 15 that more than 2/3rd CEOs believed that AI is a bigger game changer than Internet. Exhibit 16 marks only 1 in 10 CEOs are materially investing in AI.

That is a contradictory reflection of the market place. 

Why CEOs are not investing in AI? Is it because, they see it as an overhyped technology buzz that will fade away or is it because their crystal ball is not skilled enough to read AI driven implications? What is holding back strategic adoption of AI for global organisations?

In my personal opinion, at moment there is a gap between what organisations are achieving and the promise. Top three reasons are:

• organisations are failing to find right use-cases 
• treating AI projects as technology only projects
• lack of cross-functional cadence and governance

In my experience, the challenges are often not about AI technology. However, there’s a lot of noise there in terms of available AI models, frameworks, technologies, buy vs build and AI vendors. At outset, it is enticing to be drawn into a lengthy discussions on these choices. 

Recommended first approach is to start with identification of right use-cases that can deliver high value and then prioritise them by viability and deliverability of AI.

Invest in prototyping short-listed use-cases with organisation specific data to validate the promise of AI. This phase may take more time than traditional technology prototype projects. Based on the proof-points from AI results, develop indicative business cases as it will evolve over period as AI performance will improve with data and is subject to existing systems, processes and people to realise the assumed benefits.

Set a realistic expectations for out-comes and time with senior management. AI should be not used for incremental optimisations due incubation period and costs involved. For maximum benefit, AI programs demand for a major re-wiring of current ways of working, organisation and culture. Success of an AI program relies profoundly on a closer collaboration between operations, technology, commercial and management teams. A cross-functional AI team will be able to craft and deliver efficient business models, new ways of working, innovative products and new revenues.

AI projects are complex and demands for a different approach to problem definition and solving.

Just like an Agile Coach, perhaps CEOs need an AI Coach, one who can help contextualise AI strategy for an organisation’s business, data, process, technology, people and change management.