By: Ranganath Ramakrishnan, CTO Fosfor, LTIMindtree

The Greek mathematician, engineer, and inventor Archimedes said, “Give me a lever long enough and a fulcrum on which to place it, and I shall move the world.” Today, that lever of transformation is Artificial Intelligence (AI). AI and, more specifically, Generative AI can catapult enterprises ahead of their competition.

However, what will truly separate the trailblazers from the rest is the strategic application of generative AI. Our global research report, The State of Generative AI Adoption, which was recently released, shows that one-third of early adopters report a 20% or higher increase in revenue due to Gen AI, with 8% experiencing a substantial 20 to 40% decrease in costs. Notably, experienced generative AI users achieved even greater savings, with 19% reporting a 20 to 40% reduction in financial expenditures. This calls for developing the right capabilities and ensuring their adoption throughout the enterprise. At LTIMindtree, we call this AI Intensity.

Understanding AI Intensity

An organization’s ability to win with AI depends on two key dimensions – AI capability and AI adoption. The overall effectiveness of AI strategy is a direct result of these two dimensions working in harmony. Weakness in either dimension can render the entire strategy ineffective. For instance, imagine the IT department building AI capabilities, such as tools and platforms. Still, the business units are reluctant to adopt them due to a lack of clarity on how and where to utilize these tools. They are also worried about security, privacy, and governance. This ultimately leads to zero impact on the business. Similarly, executing isolated AI Proof of Concepts (PoCs) or projects without a broader strategy and capability-building results in no tangible business outcomes. In both cases, substantial investments are wasted without achieving a suitable return on investment (ROI).

Every enterprise will eventually adopt AI; the question is whether it will be driven by purpose, strategy, and intensity.

1. AI capability – It’s more than just technology

From our experience, we have observed that many organizations start their AI initiatives or PoCs by focusing solely on tool and technology selection. They identify tools, such as Feature Stores, MLOPS products, LLM Models, Vector databases, etc. Pilot projects are conducted to experiment with and learn these tools. This approach feels familiar and tangible, making it comfortable for technology teams. However, this method overlooks a crucial aspect – aligning and integrating the AI strategy with the broader business strategy, which requires more extensive work. In many cases, a well-thought-out AI strategy is completely absent.

We recommend enterprises to address the capability question across three domains holistically:

• Mindset – Generative AI is not just new; it’s different. It is necessary to get an intuitive understanding of “The Art of Possible” with generative AI. This is particularly true for senior leaders who will have to craft an AI strategy wedded to business. Investing time to learn the core concepts, strengths, weaknesses, risks, and application areas is necessary. Ensuring a good understanding of AI and its potential across the organization’s length and breadth is key to success with AI. Mindset change is key before embarking on toolset change.
• Skillset – Businesses will need new skills to thrive in the generative AI era. As intelligent automation takes over mundane cognitive tasks, qualities like curiosity, problem-finding, problem-framing, and goal-setting become crucial. Employees who can effectively collaborate with AI, enhancing their productivity, will be key to success. New roles, such as prompt engineers and prompt architects, will emerge while existing roles, like analysts and engineers, will be redefined. Every skill within the enterprise should be reconsidered and redefined to adapt to this new landscape.
• Toolset – AI also demands new tools and platforms. Generative AI introduces a new stack, ranging from infrastructure to the experience layer. We are already witnessing the rapid adoption of foundation models, LangChain frameworks, vector databases, GPU-based architectures, and more. This adds a layer of complexity to existing data analytics infrastructure. Enterprises need to strategically build their technology portfolio, aligning it with the broader strategy, to ensure a fruitful return on their investments while avoiding technical debt accumulation.

2. AI adoption – The real challenge

According to Rogers’ Theory of Diffusion of Innovations, the most challenging aspect of any innovation is its adoption. Though significant effort and cost are expended in ideation and implementation, the ultimate determinant of success lies in whether end-users and customers embrace the innovation. Without adoption, there can be no realization of benefits. AI adoption will face similar friction points and challenges, including:

• Lack of understanding of the technology.
• Absence of clear use cases that demonstrate a return on investment (ROI).
• The need to change existing ways of working for the AI strategy to succeed.
• Requirement for acquiring new skills.
• Addressing new risks associated with bias, ethics, and compliance.
• Seamless integration of AI into operational processes.

To succeed, organizations must address two key dimensions of adoption — breadth and depth.

2.1 Breadth dimension:

Like electricity and steam, AI is a powerful technology with limitless potential. It can transform every function within an enterprise. Marketing teams can leverage generative AI to create personalized content for marketing campaigns, while human resources can deploy policy chatbots for employees to access policy information easily. IT engineers can utilize co-pilots to generate code, test cases, and documentation. To truly harness the benefits of AI, applying it across the entire enterprise is crucial, creating a portfolio of use cases that can be executed over a multi-year roadmap. Every aspect of the organization, from functions to employee and customer experiences, should be transformed by AI.

2.2 Depth Dimension:

Applying AI in depth is equally important to maximize the benefits. There are generally three levels of depth to consider:

AI for intelligence generation: At this level, AI is used for standalone use cases where intelligence is generated but not fully integrated into broader workflows. For example, anomaly detection AI can identify a drop in revenue. Still, if it doesn’t provide root causes or enable action to address the issue within the business workflow, the insight will be lost.

AI for intelligent augmentation: In this scenario, AI augments and enhances a workflow intelligently. While human involvement is still required, AI is directly connected to achieving the end objective through feedback. For instance, integrating anomaly detection AI into a broader application allows business owners to take actions like changing prices or launching marketing campaigns. The application can even generate content for the campaign, ensuring that insights are transformed into action.

AI for intelligent automation: This represents the ultimate stage where AI operates in a self-sustaining manner with feedback loops. A process or workflow becomes fully instrumented, intelligent, and aware. While there is always the option for human feedback, humans are no longer required to run the process. Anomaly detection, for example, becomes powered by intelligent automation, identifying root causes and devising the right actions based on historical data and trends. It autonomously decides on price changes or selects the appropriate marketing content, executing those actions. Business owners can regularly audit and provide feedback as needed.

Historically, every new general-purpose technology, such as electricity, steam, or microprocessors, has created winners and losers in various industries. The winners are those who intentionally adopt the latest technology, building new capabilities and driving adoption across their enterprise in breadth and depth. They redefine their products and services, interact with customers in new and improved ways, reduce operational costs, and fully exploit the possibilities the new technology offers. AI, particularly generative AI, presents a similar opportunity for enterprises. The question is: Do you have the right AI Intensity to transform your enterprise and industry?

Our study distils the strategies of 450 leading decision-makers in large organizations around generative AI. You can use this study to compare and challenge your strategic choices related to generative AI and explore the value the technology can unlock.