At this stage in AI’s pervasive global growth, are the bulging seams showing signs of cracking?
Despite the daily deluge of promises about artificial intelligence, is there more to the industry assimilation of this technology in the Asia Pacific region than meets the eye?
According to a recent poll by EY (part of Ernst & Young Global Limited), mistrust, fear of potential bias (24.1%), lack of tranparency (28.6%) and explainability (22.6%) are barriers to trusted AI in the APAC region. These key factors were identified by over 70% of participants in a recent poll as the biggest barriers to raising trust levels in AI, particularly in Australia.
The poll results also showed that many believe process automation (52.3%) and generating new revenue potential through new products and processes (18.8%), are the two main benefits of AI.
Said Andy Gillard, EY Asia-Pacific Intelligent Automation Leader: “Trust is the foundation on which organizations can build stakeholder and consumer confidence and active participation with AI systems. Across the Asia-Pacific region, governments and organizations acknowledge how AI technology can deliver increasing value, in particular facial recognition technology. With the risks and impacts of AI spanning across technical, ethical and social domains, new mechanisms such as the development of frameworks and guidelines to address the unique risks of AI, are needed.”
According to Christina Larkin, EY Oceania Assurance Digital Trust Leader, business leaders must first build trust with internal and external stakeholders to clear up doubts about the data being collected and used in their AI systems. “Dealing with AI solutions that span across different jurisdictions all with different privacy regulations and cultural norms of what is socially acceptable is not an easy task.”
AI: not sure where to start?
Most organizations realize the growing value and potential of AI, which could help revive and grow multiple sectors, deliver more freedom to create, replace humans for humdrum tasks and improve customer experience and create new services. Yet, despite the clear trend and gradual increased usage, the EY poll also showed that almost half of the polling participants (41%) are not sure where to start.
Maybe this is due to a misplaced perception that AI is a solution rather than an enabler of solutions. Jeffrey Tiong, Founder and CEO of PatSnap, commented on the main way in which an AI solution differs from other conventional solutions, which is that “it is really contextually driven. People should be mindful in implementing AI to their businesses and unleash the full potential of AI depending on various company needs, purposes and circumstances.”
Suggested Gavin Seewooruttun, EY Asia-Pacific Advisory Leader for Artificial Intelligence (AI) and Analytics: “Rather than treat AI as a strictly technological effort, Asia-Pacific organizations need to view AI implementation through a human lens. To do this, leaders have to embed risk management into enablers and monitoring mechanisms for AI by demonstrating their commitment to being accountable for AI systems predictions, decisions and associated behavior. This must occur regardless of the level of autonomy for an AI system, so that ultimate responsibility and accountability for an algorithm resides with a clearly identified person or organization.”
Apart from viewing AI through human lens and recognizing it as an enabler rather than a discrete technological solution, the need to address the talent crunch for AI is also pressing. Most APAC universities have AI courses now, and consequently, people are anticipating a rich pool off talent to make it into the workforce and support organizations in leveraging this new technology. Yet, people are surprised by how diverse the specializations of that talent needs to be. This ends up in a talent crunch.
Against this backdrop, Professor Rajesh Vasa, Head of Translational Research and Development at the Applied Artificial Intelligence Institute at Deakin University, recognizes “the strong interest from governments around Asia-Pacific to develop strategies for attracting and developing talent in this space. Without such support, it may become difficult or even impossible to transform AI’s potential into reality.”
With all the trust and perception issues afoot, Gillard concluded that implementing AI in a strategic way is very important. “It is crucial for companies to rethink their approach by addressing the attributes necessary to sustain trust in their AI solutions. This will not only maximize AI’s potential, but also reassure customers to trust these technologies. In our near future, if companies employ AI wisely, consciously, and with an ethical and responsible mind, AI can make an exciting and material contribution to a better working world.”
The many provisos and caveats alluded to in the EY event constitute an important reminder from historians that technology must always be a servant and not a tool of oppression.