Don’t just learn AI, learn where it creates value: AMD India’s Jaya Jagadish on AI skilling push

India’s AI skilling push must go beyond teaching students how to use AI tools and help them understand where the technology creates value and where it does not, according to Jaya Jagadish, Country Head and SVP, Engineering, AMD India.

“You can’t just slam everything with AI because AI comes at a cost,” Jagadish told ET Online. Strong domain expertise, judgement and creativity, she said, will be critical if workers are to use AI effectively and cost-efficiently.

Her comments come as AMD and the University of Delhi on Thursday announced a strategic collaboration to train up to 10,000 students in AI skills over the next year.


Also Read: India tops global AI adoption with highest share of ‘frontier professionals’: Microsoft

Under the memorandum of understanding (MoU), students and faculty will get hands-on access to AMD’s open-source ROCm software platform for building and running AI and high-performance computing workloads on AMD GPUs, along with learning resources, technical workshops and developer GPU cloud credits.

The programme is intended to take students beyond theoretical understanding of AI and giving them practical experience in building, testing and deploying AI applications using industry-grade tools.

Jagadish said there is already a “huge rush” among students to learn AI, which she sees as a positive development. However, she said the more valuable skill in the coming years will be the ability to combine AI knowledge with expertise in a particular domain.

“AI awareness — how to build with AI and how to use AI — is, of course, very important. But along with that, your expertise in your domain is absolutely essential,” she said.

According to Jagadish, AI works best as a layer on top of existing domain knowledge, helping people improve productivity, work more efficiently and rethink how they innovate.

“You need to know where to apply it,” she said, adding that judgement and creativity would be important in determining which problems actually require AI.

The cost of AI deployment is another factor that workers and businesses will need to consider, she said.

“You need to use your judgement and ask: Does this really need AI?” Jagadish said.

Also Read: NIELIT, Intel India launch Agentic AI skilling programmes for youth across India: MeitY

From using AI to building with it

The need for practical experience is also shaping AMD’s partnership with University of Delhi.

Jagadish said classroom-based theoretical training could make students aware of AI and its potential, but they would need access to the underlying computing infrastructure to build solutions of their own.

“You need to have access to the hardware and the platform where you can build AI,” she said.

Through the programme, students will be able to use AMD’s open-source ecosystem and cloud-based GPU resources to experiment with AI applications and integrate the technology into existing processes and workflows.

Jagadish described the initiative as a platform for “fuelling innovation”, through which students can experiment, develop applications and learn from one another.

The larger objective, she said, is to democratise access to AI development and ensure the technology does not remain limited to a small group of people.

“We don't want this to become a technology that is available only to a premium class of people,” Jagadish said.

Prof Yogesh Singh, Vice-Chancellor of the University of Delhi, said in a statement that the collaboration would combine academic learning with access to modern AI computing platforms and help create a stronger pipeline of AI developers and researchers.

AI skills will need to evolve

Jagadish also expects the nature of AI-related jobs to change as adoption increases. She said AI would not simply eliminate jobs. Instead, workers would increasingly be expected to use the technology to automate certain tasks and devote more time to complex problems.

“You won't be out of a job, but your requirements will be different. Your landscape will be different,” she said.

According to Jagadish, the need for reskilling will extend beyond computer science. AI is becoming a pervasive technology that can be applied across sectors, from agriculture to space, she said.

The future talent pool will therefore need a combination of domain expertise and AI capabilities, rather than a workforce trained exclusively in AI.