Who is Prathosh AP? Why IISc Bengaluru professor's remarks on IIT students studying shastras in AI are going viral

Prathosh AP is an Assistant Professor in IISc Bengaluru’s Department of Electrical Communication Engineering. His work focuses on machine learning, deep representation learning, data-efficient learning and AI applications in healthcare. He has become prominent for linking this technical work with traditional Indian knowledge systems.

His comments followed conversations with IIT students who were studying Sanskrit and shastras with him. He questioned whether education should mean only entering a premier institution, earning a high salary and gaining material wealth. He presented traditional texts as a way to examine reasoning, responsibility and human values.

His career combines academic and industry experience. He received his PhD from IISc in 2015, worked at Xerox Research India and Philips Research, and joined IISc in September 2021. He lists 15 US patents, with 10 granted and six commercialised, and has co-founded AI-related startups.

In this article, shastras refers to traditional Indian texts and philosophical systems that examine knowledge, reasoning, duty and deeper questions about life. Their study is connected with Sanskrit and includes traditions such as Nyaya Shastra, which focuses on logic, reasoning and methods of acquiring knowledge.

Prathosh said some IIT students, including high JEE performers, approached him to study these subjects. He argued that technical achievement should not be the whole purpose of education. Students should also examine assumptions about success, material possessions and the responsibilities that accompany powerful knowledge.

He presented three dimensions of shastra study: intellectual, social and metaphysical. It could sharpen argument analysis, raise questions about responsibility in emerging technology and explore wealth, happiness and liberation. He also cautioned against studying ancient texts merely to find predictions of modern scientific discoveries.

Prathosh’s AI experience is substantial and spans research, teaching, patents and entrepreneurship. His research covers machine learning, deep representation learning, data-efficient learning, temporal data analysis and digital healthcare. He received his PhD from IISc Bengaluru in 2015 and later held academic posts at IIT Delhi and IISc.

Before returning to academia, he worked at Xerox Research India, Philips Research and a California startup. His healthcare analytics research produced intellectual property. His profile lists 15 US patents, including 10 granted patents and six that were commercialised. These figures show work extending beyond theory into applied research.

He also co-founded Cogniable, focused on learning algorithms for behavioural healthcare, and LatentForce.ai. Cogniable won first place in a Government of India AI startup challenge. He joined IIT Delhi in 2017 and IISc in September 2021, combining research leadership with teaching and company-building.

Prathosh connected AI development with human values and responsibility. His concern was not simply whether algorithms could be built, but whether people designing or controlling them understood dharma. In the article, dharma represents a framework for thinking about responsibility while technology becomes more powerful.

Using an analogy with the Mahabharata, he suggested that algorithms should be designed to improve human life. He questioned what could happen when powerful technologies are placed in the hands of people without that understanding. The article does not specify particular disasters or harms; it reports his broader warning about irresponsible technological power.

This argument gives ethical education a practical role alongside computer science. Technical progress may produce increasingly capable systems, but their goals and uses still depend on human choices. Prathosh therefore presents traditional study as one possible way to examine responsibility before, during and after AI systems are developed.

Nyaya Shastra is central to Prathosh’s intellectual argument. The article describes it as a tradition focused on logic, reasoning and methods of acquiring knowledge. He said extended study could train students to examine arguments and develop structured, analytical thinking. He also claimed that a student trained in Nyaya for 12 years could learn quantum physics and mathematics within a year, based on experiments he said he conducted.

The Upanishads support a different part of his argument. Prathosh referred to their discussions of wealth, happiness and liberation. He cited the conversation between Yajnavalkya and Maitreyi in the Brihadaranyaka Upanishad, which asks whether material wealth can lead to immortality.

Together, these examples broaden education beyond technical performance. Nyaya addresses how people reason, while the Upanishadic discussion examines what people should value. Prathosh uses both to question education focused only on prestigious institutions, lucrative jobs and possessions.

Technical education teaches students how to build models, prove results and solve computational problems. Traditional systems such as Nyaya could add another layer by focusing directly on reasoning, evidence and methods of acquiring knowledge. This matters because AI systems depend on choices about data, goals, interpretation and use, not only code.

The article gives a concrete example through Nyaya Shastra. Prathosh described it as a tradition centred on logic and argued that long study could develop structured reasoning and analytical thinking. He claimed that someone trained in it for 12 years could be prepared for quantum physics and mathematics within a year, based on experiments he said he conducted.

That claim does not mean traditional study replaces technical training. Mathematics, computer science and AI still require their own specialised knowledge. Instead, shastra study could complement them by helping students question assumptions, assess arguments and connect technical decisions with social responsibility and human values.

Machine learning is a way of building computer systems that learn patterns from data rather than following only hand-written instructions. Training examples help a system produce predictions, classifications or other outputs. Machine learning underpins much modern AI, including the research areas identified in Prathosh’s profile.

Technical methods determine how systems represent data, learn patterns and measure performance. But people choose the data, define the objective and decide what counts as an acceptable result. In healthcare, for example, an AI system’s usefulness depends on more than accuracy; its outputs must be connected to responsible decisions. The article identifies digital healthcare as one of Prathosh’s research applications.

This is why Prathosh connects AI with dharma and human values. Algorithms should improve human life, he argued, rather than exist only as technical achievements. The article presents traditional texts as a possible source of reasoning about responsibility, while technical research supplies the methods for constructing AI systems.