In this Q&A, Mohit Joshi, president at Infosys, discusses the AI investments businesses must prioritise for a solid Covid-19 recovery, those that will strengthen business, help jumpstart job creation and lead us into the 5th digital revolution.
Whilst AI has helped identify antibiotics, vaccines, and trends in the spread of Covid-19, the pandemic has appeared to decelerate the pace of AI innovation across wider industry.
With the emerging EU AI framework and a new US administration with a huge commitment to R&D investment, what can we expect from AI? Or rather, what can AI expect from us?
Joshi helps Information Age explore how AI can lead businesses to a strong recovery, the huge opportunities to business if they can get it right and most importantly, how to get it right.
What AI investments should businesses prioritise for Covid-19 recovery? Which ones will strengthen business, lead to job creation and lead us into the 5th digital revolution?
In industries that are traditionally dependent on physical interactions, Covid-19 has accelerated remote working and digital transformation. For some this has meant a sudden overhaul of processes and technology, and the acceleration of this change will enable them to be more agile and intelligent in the future. Of course, this will not come without challenges particularly in terms of personnel and skills. However, by enabling and converging AI with technology such as IoT, businesses can unlock the ability to continue the smooth running of operations while maintaining physical distance. We can expect this to really drive value for manufacturers, retailers, and logistics as well as infrastructure maintenance providers, and even airports and hospitals way into the future.
And, while there will be a whole host of systems that businesses and individuals will interact with in the future, they must be intelligent, they need to involve us, they need to sense and be able take decisions, some on their own. What this means for businesses is that while the digital presence of systems and processes will only increase, increasing their intelligence and continually enhancing them will be crucial. Therefore, we can expect the role of AI to be far more strategic than ever before, particularly as we think about emotional intelligence in the future. The beauty of this change will be greater demand for people and skills. While AI will start making systems intelligent and reduce demand on maintenance and smaller operations, the next innovations, the roadmap development, the enhancements, and emotional intelligence will require more man=power.
Up to now AI investment in industry has been aimed at solving specific business challenges and driving cost reduction, now businesses really need to invest in creating an enterprise grade AI stack to responsibly scale AI across the enterprise. Ultimately, organisations need to focus on improving the end user and customer experience, using AI to drive hyper-personalisation such as conversational commerce tools. Service based industries, those that involve these consumer interactions, are where we see the greatest opportunity for AI to strengthen business.
2. What is the 5th digital revolution?
We often hear that we are living through the fourth revolution right now. A time where we have seen the development of enterprise-grade cloud computing, which has made digital power and agility available to everyone. However, as we move in to the fifth digital revolution, it’s AI that will be the driver. In fact, Gartner predicts that by 2022, the business value created by AI will reach $3.9 trillion – and that’s just the beginning of the revolution.
AI investment to increase but challenges remain around delivering ROI
3. Why is explainable AI the cornerstone of successful AI investment and adoption?
Explainability of AI solutions is increasingly receiving focus at the moment, particularly in the face of rapidly developing systems, regulatory attention and above all the growing importance of trust for businesses and consumers alike.
As businesses and consumers, but also official bodies and regulators become more dependent on AI-based systems, being able to explain the chain of reasoning that led a system to a prediction to demonstrate clear accountability will be critical to ensuring trust and transparency. We’ve seen mechanisms put in place to tackle the potential problems stemming from the rising importance of algorithms such as the ‘right to explanation’ in the GDPR, and in the future we could expect Explainability by Design as standard. Organisation’s governance will also need to reflect that approach.
4. How can organisations get governance right with AI investment?
It is true that AI is enterprise ready, but many enterprises are not ready for AI. First and foremost, organisations need to address the challenges of governance to effectively adopt AI, adhere to regulatory pressures and reap the technology’s vast benefits in a trustworthy manner.
The key challenge we need to overcome is that solutions driven by AI and Machine learning enabled systems largely depend on the training data set, and data sets could be biased. Without the right governance, these issues could cause reputational damage and regulatory compromise. Organisations that do not deploy AI ethically will soon run into serious trouble with regulators, and a strong ethics framework should be the very foundation. If organisations can get this groundwork in place and right now, they will have competitive advantage when industry leading regulation such as the EU’s AI framework is adopted, already operating with the technology in a trust-worthy and responsible manner.
What ’s more, accountability, fairness and explainability should continue to be the guiding principles for any progressing organisation employing AI. Investing in creating processes, metrics and an enterprise AI stack will be critical to ensuring AI is measurable and deployed accountably.
But we have to be clear. We are all still learning what best governance should look like as innovation in the AI space is happening at such a rate that even experts are struggling to keep pace. We will constantly have to update both our ethics and governance, to be ready for every evolution in AI technology, observing the machines to figure out what they do, and to ensure they are not used wrongly.
