The Horse, Humanity and the AI Opportunity: Translating Technological Capability into Economic and Social value

There is considerable potential for AI to create employment in the medium and long term. The opportunities will not be confined to the companies that manufacture chips or develop LLMs. They can extend to data preparation, software development, integration, consulting, cybersecurity, maintenance, specialised applications, training and a wide range of supporting services.

Image
Representational Photo

The horse has captured the imagination of humankind since time immemorial. From Indra’s divine horse, Uchchaihshravas, to the four white horses drawing Arjuna’s chariot on the battlefield of the Mahabharata, with Lord Krishna holding the reins; from Pegasus and Arion in Greek mythology to Chetak, the legendary horse of Maharana Pratap, horses have inspired romance, courage, loyalty and adventure across cultures and generations.

Over millennia, horses have served a remarkable variety of purposes. They have carried warriors into battle, drawn chariots and carts, transported travellers and supported agriculture. Who can forget the horses of the Wild West, ridden by cowboys to herd cattle across vast open ranges? Today, horses are also associated with show jumping, dressage, other equestrian sports and even the simple pleasure of a ride along the beach, delighting children and adults alike.

Horses may belong to the same species, but their breeding, physical attributes and training vary according to the roles they are expected to perform. Professional breeders select and develop horses for particular characteristics, while trainers prepare them for specific tasks. Their eventual performance depends not merely on their breeding, but also on the quality of their training, the skill of their riders and the environment in which they operate.

This brings me to the analogy I wish to draw with artificial intelligence. AI is not just about chips and large language models (LLMs). It is about the entire ecosystem that develops the technology, adapts it to different purposes, trains people to use it, and enables them to harness its capabilities effectively.

The Many Components 

Chip design is a highly specialised activity requiring advanced engineering skills, research and substantial investment. India already has a significant presence in semiconductor design, including through Global Capability Centres (GCCs) established by multinational companies. This provides a foundation on which India can build greater capabilities in chip architecture, design tools, verification and specialised processors for AI applications.

Chip Manufacturing

Designing a chip and manufacturing it are two different capabilities. Taiwan is a major centre for advanced semiconductor manufacturing, while China has built substantial semiconductor manufacturing capacity. The Netherlands occupies a critical position through companies that manufacture advanced lithography equipment required for producing leading-edge chips.

India is developing its own semiconductor manufacturing ecosystem. This is a long-term undertaking involving capital, technology, skilled personnel, equipment, materials and a reliable supply chain. It would be unrealistic to expect a country to acquire all these capabilities overnight.

Building Large-Language Models

LLMs form an important part of the AI ecosystem, but not every application requires the largest or most expensive model. General-purpose models can serve a wide range of tasks, while smaller or specialised models may be more suitable for particular industries, professions and local-language requirements.

India's opportunity lies not merely in developing indigenous foundation models, but also in building models and applications that address its own requirements: Indian languages, agriculture, healthcare, education, law, manufacturing and public administration. The choice of model should be determined by the problem to be solved, its cost and the quality of the results.

Data Creation, Management and Storage

AI depends on data, but the quantity of data alone does not determine its usefulness. Data must be accurate, relevant, well-organised and suitable for the intended purpose. In many applications, collecting, cleaning, labelling and maintaining data can be as important as developing the model itself.

Data centres provide the computing and storage infrastructure required to process this information. Their development, in turn, creates demand for construction, networking, equipment, security, maintenance and specialised technical services.

Supporting Infrastructure

AI requires reliable computing infrastructure, electricity, cooling systems and high-speed connectivity. Large data centres can consume substantial amounts of power, making energy availability, efficiency and cost important considerations.

The ecosystem therefore extends into power generation, transmission, cooling technologies, networking, equipment manufacturing and infrastructure management. These are not always visible when we discuss AI, but they are indispensable to its operation at scale.

General-Purpose Applications

AI is increasingly being incorporated into familiar productivity tools such as Microsoft Office. Drafting documents, analysing spreadsheets, preparing presentations, summarising information and automating routine tasks are becoming part of the everyday use of AI.

The economic value here does not depend on every user building a model. It depends on how effectively people use these tools to improve productivity, save time and perform tasks that were previously difficult or expensive.

Enterprise Applications and Integration

Platforms such as Oracle and SAP support complex business operations across finance, manufacturing, logistics, human resources and supply chains. Integrating AI into these systems requires much more than simply adding a chatbot.

Existing databases, software, business processes and organisational requirements must be understood and adapted. AI solutions must be tested, integrated and customised to meet the needs of individual organisations.

This creates opportunities for IT services companies, consultants, startups and GCCs. India has considerable experience in enterprise software and technology services, and this experience can be extended to AI implementation.

Maintenance, Security and Seliability

AI systems are not developed once and then left to operate indefinitely. They require monitoring, updates, troubleshooting, performance evaluation and continuous improvement.

They must also be protected against cyberattacks, misuse, data leakage and unauthorised access. Their outputs need to be checked for accuracy, particularly in applications where mistakes can have serious consequences.

As AI becomes embedded in business and public services, cybersecurity, privacy, reliability and responsible oversight will become integral parts of the ecosystem. These activities require skilled people even when the systems themselves become increasingly automated.

Specialised Applications 

AI is not confined to computers, office applications and data centres. It is also finding applications in industrial machinery, robotics, drones, vehicles, medical equipment, agriculture and defence.

Such applications may require AI to operate alongside sensors, cameras, communication systems and physical machinery. Some will depend on large cloud-based models; others will use smaller models operating directly on local devices, a field commonly known as edge AI.

The development, installation, testing, repair and integration of these systems open up opportunities across engineering, manufacturing and technical services.

Breeding the Horse 

The ecosystem described above is much larger than we often visualise. Its various components are interconnected, and each creates requirements for specialised skills, infrastructure and services. Although AI can increasingly automate parts of this work through self-learning and other capabilities, the need for human expertise does not simply disappear.

Consider the difference between developing an AI model and applying it to a real-world problem. A doctor must understand medicine to judge whether an AI-generated suggestion is clinically useful. An engineer must understand the underlying system to determine whether a proposed solution will work. An accountant must know the principles of accounting to identify errors in an AI-generated analysis.

AI can assist these professionals, sometimes substantially, but domain knowledge remains important for framing the right questions, evaluating results and accepting responsibility for decisions.

This is where the horse analogy becomes particularly relevant. Breeding a superior horse is one thing; training it for a specific purpose is another. The rider must understand the animal, the task and the conditions in which it will operate. The same principle applies to AI.

The ability to develop AI is one skill; the ability to apply it productively in a particular field is another. A country needs both, but the number of people who can use AI effectively across different occupations may be far greater than the number involved in developing foundation models.

For India, this distinction is important. We need highly skilled researchers and engineers, but we also need teachers, doctors, accountants, lawyers, technicians, administrators, farmers and entrepreneurs who can use AI to improve their work. We need people who can build applications, integrate systems, maintain infrastructure and adapt technology to local requirements.

Employment Outcome

There is considerable potential for AI to create employment in the medium and long term. The opportunities will not be confined to the companies that manufacture chips or develop LLMs. They can extend to data preparation, software development, integration, consulting, cybersecurity, maintenance, specialised applications, training and a wide range of supporting services.

At the same time, we should not assume that every new AI application will create more jobs than it displaces. AI can automate existing tasks, reduce demand for certain roles and change the nature of many occupations. Some new jobs will require skills that displaced workers may not possess.

The ultimate employment outcome will depend on investment, the pace of adoption, the creation of new products and services, the expansion of markets and the ability of people to acquire new skills.

For India, the challenge is therefore not simply to adopt AI or compete in building the largest models. It is to develop the capabilities that enable Indian businesses, professionals and workers to participate across the value chain. This requires education and training, industry participation, research, entrepreneurship and the willingness to adapt established ways of working.

Developing the People 

The horse has served humanity for thousands of years, but its usefulness has always depended on the purpose for which it was bred, the training it received and the person who guided it. A horse trained for racing is not necessarily the best choice for pulling a cart, just as a horse suited to agricultural work may not be suited to show jumping.

AI is no different in principle. A powerful general-purpose model may be impressive, but a smaller, specialised system that solves a particular problem reliably and economically may deliver far greater value in its intended setting.

The real AI opportunity lies not merely in building the horse, but in training it, learning to ride it and putting it to the right use.

For countries such as India with a technological and human resource base, the objective should be to participate across the entire AI ecosystem - from chip design and manufacturing to data, models, infrastructure, applications, integration and maintenance. Equally important is developing the people who can use these capabilities productively.

The countries that benefit most from AI will not necessarily be those that develop only the most powerful models. They will be those that build the skills, enterprises and institutions capable of translating technological capability into widespread economic and social value.

(The author is an Indian Army veteran and a contemporary affairs commentator. The views expressed are personal. He can be reached at  kl.viswanathan@gmail.com )

Post a Comment

The content of this field is kept private and will not be shown publicly.

Newsletter Subscription

The subscriber's email address.
Stay informed - subscribe to our newsletter.