In the short story “The Library of Babel”, Jorge Luis Borges writes of an infinite library filled with books that are believed to contain every possible permutation of the alphabet. Some inhabitants of the library devote their lives to scouring its shelves, in search of a “perfect” book which can catalogue the universe.
Perhaps their job would be made easier with AI. The tech has the potential to drastically expand the ways governments process and use data, and is already making waves across Asia Pacific.
This is what we discovered in the recently released IDC InfoBrief, ‘Artificial Intelligence Practices That Deliver Results’, commissioned by leading AI and machine learning firm Dataiku. We delve into the latest insights on how companies in Asia Pacific are adopting AI, and how the trends compare with 2020’s findings.
Ready for an AI boom
The economic trouble brought about by the pandemic has done little to deter AI spending in 2020, IDC found. A majority of firms reported no changes – some even increased investment in AI.
Organisations have instead shifted their focus towards being future-ready with AI. Around twice as many countries have started investing in AI tech during 2020 as compared to 2019, while those that have already invested in AI have further matured in their business processes.
Organisations have tasted the fruit of success with greater AI maturity, addressing more complex use cases. However, we still see that most models still remain in pre-production stages – companies are still missing a few vital ingredients that prevent them from demonstrating the value of AI.
4 key ingredients for AI adoption
In order to successfully scale AI, companies need to balance resources and efforts across four key areas.
First, organisations should have the appropriate tech capabilities in place. IDC recommends using a central AI platform – which will allow flexible development of solutions as well as universal access to intelligence services.
But forget the tech – IDC shows how data, people and process are woven into every stage of the maturity curve and are core to effectively scale their AI initiatives.
Second, companies need to improve data readability. This may include working on data acquisition, access, preparation, security, governance, integrity, and quality. A lot still has to be done: APAC organisations investing in AI barely pass data readiness measures.
Third, staff need to be equipped with relevant skills to boost AI innovation, while promoting strong collaboration across teams. This can be done with regular upskilling programmes as well as good talent management. Companies are restructuring into hub-and-spoke models, in which a central ‘hub’ leads a network of distributed ‘spokes’, rather than concentrating efforts in Centres of Excellence.
Last but not least, one of the most important trends highlighted by this InfoBrief is the increasing importance for companies to systematically execute AI, while continuously improving them. More models in production means more complexity. This leads to the need to standardise and monitor data and models, so firms can mitigate associated risks . IDC highlights also the need for process metrics in order to measure (and increase) business value and create momentum.
Borges’ story may prize a “perfect” book that catalogues the universe. Governments in the region are after something of matching value: seamless digital citizen services that would change lives. AI could be a crucial tool.
For additional insights, read the full IDC InfoBrief Artificial Intelligence Practices That Deliver Results here.