AI STRATEGY says Azfar.“ We knew we had the data and the data science, and machine learning capabilities were maturing, so the focus was on how we could take data from operations, networks and infrastructure to understand investment priorities based on real, measurable information.
“ The notion of prioritising investment was a big aha moment for the entire industry,” he adds.“ It suddenly moved everyone from a mindset of‘ let’ s invest everywhere’ to‘ let’ s use intelligence to prioritise customers, domains, services or areas very effectively to achieve a better ROI’. We ultimately created the world’ s first AI capability for taking that data and using it to inform investment strategy in the telecommunications industry.”
After a period of trialling, testing and development with Bell Labs’ research teams( Nokia ultimately acquired Alcatel-Lucent in 2016), largely to build trust and confidence in then-fledgling AI technologies, the technology was ready.“ We knew we were sitting on some very powerful capabilities,” Azfar recalls.“ There was a journey of transition into adoption but, by the time I left the organisation, 20 or 30 customers worldwide had started implementing those AI capabilities.”
How are you operationalising AI? AI adoption is continuing at pace across all industries and sectors. According to McKinsey’ s Global State of AI Survey, for example, more than 75 % of organisations now use AI in at least one business function.
NOKIA: SIX PILLARS FOR RESPONSIBLE AI ADOPTION
1. Fairness
2. Reliability, safety and security
3. Privacy
4. Transparency
5. Sustainability
6. Accountability
The consulting firm also found that as many as 88 % of large enterprises are undergoing an AI transformation.
However, sit that impressive growth next to research from Accenture that finds 76 % of C-suite execs admit to struggling to scale AI beyond pilot projects, and Azfar’ s focus on value and tangible outcomes remain as pertinent as ever.
“ You have to be pragmatic,” he says.“ Start by considering the real business problem to solve, and also ask whether it’ s worth solving. Are you creating new products and services to generate revenues, or are you applying automation to help reduce CAPEX? Whatever area you’ re focused on, you need to ask if the problem can be solved – all technologies have their limitations, after all. If you don’ t put AI or any other tech through the business processes, it just ends up as an R & D project. aimagazine. com 49