Recommendations of the 2019 Asia Pacific AI Readiness Index

This article outlines the key recommendations of the Asia Pacific AI Readiness Index. Part 1 of the article provided an overview of the report, in terms of the rankings, the criteria used and overall trends observed. Part 2 detailed the key findings of the report.
 

1. Preparation of AI talent: There are only 300,000 AI engineers, researchers and practitioners worldwide, when several millions of them will be needed over the next two decades.
   • Adapt educational institutions and what they teach:Singapore has introduced coding to primary and secondary school curricula. AI-specific initiatives need to be developed around more abstract facets of AI i.e. role of ethics in computer science, influence of biases in society and databases and importance of data quality in technologies. For use across all complex AI applications, AI curricula must move beyond a purely technical focus, to include perspectives from history, sociology, ethics, philosophy etc.
   • Use AI to make education better: Big data and analytics can help improve teaching and learning experiences by capturing data on student demographics, school attributes, individual trajectories etc. for resource allocation, policy adjustments and improving workflows.
   • Support upskilling and lifelong learning schemes: The effectiveness of AI policies depends upon the capitalization of AI-specific advanced skills and knowledge to develop and maintain complex systems and technologies. It is estimated that 14% of the workforce is likely to make a transition to new occupational categories.
   • Build a diverse and representative workforce: New and transformative AI technologies must effectively address the needs and expectations of diverse populations. Thus, the creators must also be diverse i.e. people of all backgrounds must produce the technological innovations. Increasing diversity also has the advantage of increasing the pool of AI-skilled talent in the long run.
2. Build trust in AI:Privacy and security are central to the development and deployment of AI technologies. Thus, algorithms must be understandable, transparent and trustworthy, with accountability of the organisations using them, to avoid errors and breaches. How can consumers and businesses be reconciled with the idea of letting machines influence important aspects of their lives?
   • Adherence to privacy principles and practices: “Privacy by Design” comprises numerous technical, organizational and security measures at each stage of developing data-dependent products, systems and operations. Conducting regular privacy impact assessments can identify and mitigate security risks before actual processing of personal data.
   • Making transparency a key feature of AI products and services: The complexity of digital products and services entails complex uses of data. Thus, systems must be transparent in terms of not only how, but why a decision on an action was reached. This will allow consumers to better understand how AI systems work and affect them.
   • Controlling data collection and usage: Data anonymization refers to the deletion or pseudonymization of personal data to make it irreversibly untraceable and unreproducible, thereby, allowing AI systems to use data without divulging private information and minimizing risks of breaches or accidental disclosures. Data limitation policies force organisations to limit the use of personal data for the purpose it was originally collected for. Data minimization reduces the amount of data collected and processed by establishing at the onset what data is relevant for a specific purpose.
   • Enable ethics in AI: For AI to grow and deliver on its promises, AI must be developed, implemented and monitored following the highest ethical standards. Policy-makers and professionals must ensure that fairness and diversity is built into the datasets because AI algorithms are only as good as the data they are fed to learn, in making recommendations and predictions.
   • Using AI to strengthen cybersecurity: AI can improve existing detection and response capabilities and create new preventative defence protocols. It allows the identification of threats sooner by sifting through large volumes of data and evaluate patterns faster. It can also help streamline complex, manual or time-consuming processes.
3. Shape AI ecosystems: Governments drive the adoption of AI. The public sector can influence the expansion of AI technologies across investment and procurement schemes, education, labour, migration policies etc. Government-led initiatives must enable the growth of the AI value chain. How can governments effectively foster dynamic, sustainable and innovative AI ecosystems?
   • Make data open and available: AI technologies require a secure and steady diet of reliable data to learn and function. Major data gaps in relation to the existence of fragmented data or collected data that has not been analysed, must be fixed by making non-sensitive data freely available for access and use by everyone. This approach can help the implementation of innovative solutions to truly address societal needs and expectations.
   • Facilitate cross-border flows and regional cooperation: Data flows across multiple jurisdictions must be secured with the consistent cooperation between governments as people, devices and platforms remain constantly connected beyond geographical borders.
   • Set an example by using AI in government organisations: Modernizing and improving the public sector can eliminate administrative processes, support overburdened institutions, optimize resource allocations and prevent crime. AI adoption by the government cannot follow the same historical arc of technological adoption i.e. at a pace and scale lesser than the private sector.
   • Create specific rules to enable ethical and humane uses of AI: Many economies have created new roles within specialized government agencies to design unprecedented solutions to unprecedented challenges in a digital era, especially related to the legal, moral and ethical dilemmas/issues created by AI technologies. Governments must be committed to keeping AI safe and ethical for consumers and businesses alike.
   • Use AI for good and teach AI to do good: According to the AI for Good Global Summit, AI must help humanity solve grand challenges by capitalizing upon unprecedented quantities of data generated on health, commerce, communication, migration etc. In collaboration with private sector entities, governments must share AI tools and resources, datasets, knowledge and expertise to address sustainability challenges and help vulnerable populations.

 
India Outbound
April 12, 2019

 
 

source https://indiaoutbound.org/recommendations-of-the-2019-asia-pacific-ai-readiness-index/