Sustainable AI: Challenge Of Data Integration, Digital Divide And Investment In Advanced Technology
Systems that use artificial intelligence (AI) are already revolutionising industries and ingraining themselves into every aspect of our everyday lives. These systems, which employ machines to analyse massive amounts of data, have fundamentally altered how people work and play and are currently used in a wide range of industries, including banking, energy, and agriculture.
“AI can be used to predict rainfall, and weather pattern changes- not just for the day, but also for an entire year. Many of the things are in the making and testing process. We have to make sure to use it positively. With such an analytical and predictive structure, it could be a very good planning tool for the government and public policymakers,” said Chaitanya Kalia, Partner and Climate Change and Sustainability Services Leader, EY India. Precision in agriculture, sustainable supply chains, environmental monitoring and compliance, and improved weather and disaster prediction are just a few examples of the areas in which AI is being incorporated.
Data plays a very essential role in training and upgrading AI. But it comes with an environmental cost. According to the Massachusetts Institute of Technology (MIT), the carbon footprint of the cloud has surpassed that of the whole airline sector, and one data centre may use as much electricity as 50,000 households. In addition, the datasets required to train AI are getting bigger and consume a lot of energy.
“AI has revolutionized every aspect of our lives. Data is core to all AI solutions. Traditionally, it wasn't accessible to wider organisations across the globe. The fundamental shift that has happened in past years is the general availability of data, with its own advantages and disadvantages. But, in general availability of data has opened up a lot of avenues,” said Piyush Saxena, SVP, Cloud Business and Offering Management, HCLTech.
A 2019 report by MIT also suggested that training just one artificial intelligence model can result in emissions of more than 626,00 pounds of carbon dioxide equivalent, or about five times the lifetime emissions of a typical American car. For example, according to studies, Google's AlphaGo Zero, an AI that competes with itself to learn the game of Go, produced 96 tonnes of carbon dioxide during its 40-day research training period. The carbon footprint of ten typical Singaporeans over a year is comparable to that, as is 1,000 hours of air travel.
“We handle almost 5 billion transactions per day. Back in 2010, we converted our data centre to be more energy-efficient. In 2019, we reached the goal of being a carbon-neutral data centre. In this process we have observed the shift from cost-centric behaviour to more carbon-sensitive options,” told Nagesh Kartheek, Director of Engineering, Amadeus.
Integrating human abilities into AI solutions is the biggest and most underutilised possibility to assist in closing the digital divide. Like many other innovations, AI is probably going to split the salary distribution in two. As workers become more productive performing these new, more complicated AI-assisted jobs, their position in the global wage distribution is likely to improve. The digital gap between Indian men and women remains pronounced despite fast improvements in internet access, particularly via mobile internet use, which nearly doubled from 2018 to 2020. The National Family Health Survey-5 analysed how often men and women in India use the internet. They discovered that only 33% of women have ever used the internet, compared to 57% of males.
“A very notable thing about AI is unlike other technology which started with higher margin businesses and then trickled down to the affordable segment, there is no trickle down in AI. To serve those who are at the bottom of the pyramid you need innovators who can go into spaces where there is very little money to be made, at least at the beginning,” said Sudha Srinivasan, CEO, The/Nudge Centre for Social Innovation.
AI might be an effective way to open doors for the corporate world of the future. Numerous corporate processes will eventually be automated using AI. International Data Centre (IDC) projects that global spending on AI will double between 2020 and 2024, rising from USD 50 billion in 2020 to USD 110 billion in 2024. It also observed that the growing use of AI in finance is made possible by the abundance of data available as well as improved, more inexpensive computer power, which is used in fields including asset management, algorithmic trading, credit underwriting, and blockchain-based financial services. “There is a new realisation that sustainability is a new class of risk. Many banking professionals have spoken about how they think of sustainability as an investment. Assess and report, transition to a more equitable future, and building agility and resilience are three broad areas to work upon,” said Sathish Seshadri, Global ESG Leader, Genpact.
