How to successfully use artificial intelligence in our businesses

Generative artificial intelligence (AI) has created much excitement about the potential of AI applications that is here to stay, and could very well dramatically transform how we live, work and interact. While it is great to weave a poem, paint a picture and generate a report or code, the real challenge is how businesses can leverage this technology to build applications that boost employee productivity and/or improve user experience. An even bigger question is: how can a corporation adopt this technology at scale for its business needs with trust, transparency and confidence in its outcome?

Over the years, AI has steadily made it to the centre of driving digital change across industries. The International Data Corporation's (IDC) February 2023 Worldwide Artificial Intelligence Spending Guide forecasts AI spending of $3.6 billion by 2026 in India, with an expected compound annual growth rate (CAGR) of 34.3%. Generative AI has revved up the imagination of not just tech startups but a host of organisations, which are building large language models (LLMs) or use cases for their business needs.

From improving manufacturing, sales functions, marketing content and customer service to experimenting with non-English Indian language use cases for rural populations, the technology offers infinite potential. And India Inc is not losing any time tapping into it.

From supporting individual employee efforts to elevating the productivity levels of the entire organisation, the technology can drive significant improvement in critical workflows. While text summarisation, semantic search, content creation and code creation are the most popular, organisations can lay hands on customer insights in near-real time, complete internal audits faster, improve employee satisfaction and customer trust, and drive business growth.

But these use cases are only the beginning. Generative AI can prove immensely powerful in maximising user experiences and augmenting decision-making capabilities in the future. Therefore, it's important for us to start preparing the workforce to work with AI, synergising their knowledge and expertise for better and faster outcomes. Generative AI will not only accelerate AI adoption but also scale the impact of AI.

Foundation models are the foundation of generative AI: Large foundation models that have been trained on large amounts of data form the basis of generative AI. Most of this data is unlabelled or unorganised. They serve as the starting point for the development of more models that can be tuned to enable a specific use case or domain. Organisations must invest in building foundation models that have been trained on their proprietary data and controlled external data sources, to avoid risks related to intellectual property (IP) or copyright issues, as well as to tackle trust, transparency and privacy issues.

Open platform for seamless integration: As generative AI presents an opportunity for value creation, to be innovative, organisations will use not one but many models. Considering the costs attached to training and maintaining these models, organisations can base their decisions on cost, effort, data privacy, IP and security. Organisations must also adopt an open platform approach, such that they benefit from models available in the open. Since these models will be spread across their hybrid cloud environments, they will need a platform that integrates all these models and deploys them for their use cases and builds value for their organisation.

Governance framework to build trust: The most important element is building a governance framework that ensures trust, accuracy and confidence, and adheres to regulatory requirements. Undeniably, generative AI and LLMs open up tremendous possibilities for organisations to transform and grow.

But they also introduce hazards and risks in the process related to biased or profane language and personally identifiable information. It's important for organisations not to get swept away, but that they anchor themselves with ethics and governance mechanisms to make the exploration of AI an enduring journey.

The success of AI is not 'more of it' but effective adoption with thorough understanding, planning and deployment for speed and scale. And, given its incredible benefits, it should have guard rails in place to ensure it is effective and a force for good - for both society and economy.

(The writer is managing director, IBM India)