When ChatGPT burst on the scene in November 2022, with 100 million users in the first two months, Generative AI went from niche to mainstream. That said, there is still a good deal of confusion about what generative AI is, what it does, and what it all means. In addition, the technology is evolving rapidly.
To start with the basics, gen AI refers to computer models that use artificial neural networks to mine data sets and generate content, including text (ChatGPT), images (Dall-E, Stable Diffusion), and audio (Aipa, Deepgram, Assembly AI). Gen AI has been used to automate tasks, such as simple customer-service conversations; to accelerate work, such as document summarization; and to augment human capabilities, such as software coding. From banking and pharma to retail, fashion and (of course) tech, interest and investment in it are growing fast. And its potential is enormous: McKinsey has estimated that gen AI could add $2.6 trillion to $4.4 trillion annually to the global economy in the form of increased productivity.
What does this mean for India? Given our already strong position as a technology services provider, excelling at gen AI could open a new frontier in innovation and entrepreneurship. Overall, we believe that it could add more than $100 billion to the value of the Indian economy.
Given the breadth of gen AI’s possible applications, businesses that want to get started should establish cross-functional groups to coordinate their approach. These teams would be responsible for training, technology development, and establishing new partnerships if necessary. Most important, they need to balance the risks of gen AI with the value to be created.
There are several ways that India can profitably use gen AI right. Here are five of them.
Sales productivity: While digital channels are growing fast, people-driven channels such as call centres and stores are still dominant in India. There is valuable consumer data embedded in e-mails, chats, images, call recordings and other sources, but conventional machine learning models have not explored them. Gen AI tools can sift and deploy this data in new ways. Equipped with easy access to information on a product portfolio, for example, sales staff can offer intelligent nudges to consumers while managers can improve their coaching. This is relevant to a wide range of sectors, including consumer products, retail, banking, insurance, automotive, building materials and pharmaceuticals, among others.
In a similar manner, gen AI tools can help business-to-business (B2B) sales teams improve their productivity by helping them review proposals, write new ones and note areas that need attention.
Customer engagement: By enabling digital self-service and augmenting agent skills, gen AI can improve productivity in this field by 30-45%. For example, it can help get less-experienced agents up to speed faster, and thus deliver strong customer service. In addition, gen AI tools can provide a comprehensive overview of past customer interactions; real-time assessments of customer sentiment; and personalized recommendations. It can also solve customer problems and indicate where communications can be targeted. And it can do all this quickly. The likely result is greater customer loyalty.
Technology services: Over the past decade, automation productivity has risen 8-10% a year in India. Gen AI could do even better, faster. Gen AI can handle multiple programming languages; the tools are also easy to deploy, requiring limited setup and training, and are intuitive to use. In controlled pilot tests, gen AI tech tools have delivered 25-50% productivity improvement in development, documentation and testing. The potential is therefore significant for digital ventures, software-as-a service companies, and in-house software development teams. Finally, gen AI can help define novel offerings for businesses and customers.
Business functions: Gen AI can support functions such as finance, personnel, procurement, tech support and legal by answering queries based on existing policies, summarizing documents, generating drafts and reviewing contracts. Early results from such projects indicate the possibility of productivity improvements of 20-30%.
Learning effectiveness: Gen AI tools can provide coaching support to learners from kindergarten into the workplace, whether by offering content, conducting evaluations or making tailored recommendations for further learning. In addition to technical skills, it can enable tools for soft skills training, such as communications, that can get early tenure and low-skilled workers to perform better, faster. It could also revolutionize schooling in settings where there are not enough quality teachers; gen AI tools could act as virtual assistants that support students and notify human teachers when they need attention. In healthcare, they could play a similar role.
In short, gen AI will likely find applications in all domains and unlock significant new opportunities. However, there will also be challenges. Gen AI models are prone to ‘hallucinations’—providing responses that can sound plausible but are wrong. Security, as well as data and intellectual property protection, need to be addressed. Financial costs of usage and environmental costs of development could be high. As with any other new development, gen AI needs to be managed with organizations systems and a culture for digital trust and responsible use of AI.
There is also concern that gen AI’s ability to improve productivity could affect employment in sectors like software development and call centres. Large-scale adoption of gen AI for value creation is more likely in organizations and societies that train people in new skills and then match them with new roles, including in areas that emerge from gen AI applications.
India is a global leader in technology services and entrepreneurship. Leading gen AI adoption offers a path to strengthen its position and create new opportunities.
Rajat Dhawan, Sathya Prathipati and Ankur Puri are, respectively, a managing partner, senior partner and partner at McKinsey & Company India.