Keynote-Vishal Sikka
19 Feb 2026 12:00h - 12:15h
Keynote-Vishal Sikka
Summary
The session began with Speaker 1 thanking Sir Hasabis and introducing Vishal Sikka, founder and CEO of VNI and former Infosys CEO, as a leading thinker at the intersection of AI and enterprise [1-10]. Sikka opened by noting that users who truly understand how to employ generative AI can achieve dramatic productivity gains, citing a former classmate who rebuilt a nine-month, 15-person service in just 14 days using an AI coding tool-a more than 250-fold improvement [15-20]. He further illustrated the impact with a customer in the home-goods distribution sector who, using his company’s AI product, reduced a decision that would have taken a year to a few days, highlighting AI’s power to provide instant, multilingual knowledge synthesis [20-22]. Sikka emphasized that such effectiveness is uneven, describing it as a “jagged frontier” where only those who grasp AI’s capabilities can reap its benefits [24-26].
His second point stressed that mastering AI requires not only technical knowledge but also awareness of its limits and the ability to bridge the gap between large language models and enterprise users through reliable, verifiable systems [27-31]. He argued that closing this gap can transform legacy systems, simplify complex enterprise processes, and empower end users, a vision supported by remarks from Mukesh Bhai earlier in the summit [33-38]. Sikka highlighted India’s abundant talent and the Prime Minister’s call for a billion entrepreneurs who can create value with AI, positioning the country as poised to lead this transformation [40-43].
The third point warned that today’s AI still suffers from serious shortcomings such as hallucinations, limited physical understanding, and safety risks, which must be addressed before broader enterprise adoption [44-56]. He compared AI’s energy demands to a 100-watt light bulb, noting that current models consume far more power than the human brain and that substantial efficiency gains are still required [62-66]. Sikka invoked historical Indian innovations-the Green Revolution and mass connectivity through Jio and Airtel-to illustrate how the nation has repeatedly leapfrogged challenges and can do so again with AI [70-73].
He concluded that the summit demonstrates a path toward a “human revolution” powered by purposeful, safe AI, where billions of entrepreneurs not only earn livelihoods but also enrich lives [74-75]. The discussion underscored both the extraordinary opportunities AI offers for productivity and societal impact, and the urgent need for responsible development, safety, and energy efficiency to realize those benefits [22-56]. Ultimately, Sikka’s message positioned AI as a catalyst for India’s next wave of innovation, calling for coordinated effort to master, improve, and responsibly deploy the technology for broader human benefit [68-75].
Keypoints
Major discussion points
– AI can deliver dramatic productivity gains for skilled users.
Sikka cites a Stanford-classmate who rebuilt a nine-month, 15-person service in just 14 days with a generative-coding tool – a “more than 250 times improvement in productivity” – and a home-goods distributor who reduced a year-long country-exit analysis to a few days using AI-driven simulations. [15-20]
– Bridging the gap between large language models and enterprise users is essential.
He stresses that effectiveness with AI requires not only technical know-how but also awareness of limitations; enterprises need “trusted, verifiable, reliable systems” that sit above raw LLMs to deliver correct business value. [27-32]
– Current AI systems have critical limitations that must be solved before wide-scale adoption.
Sikka highlights hallucinations, safety risks (e.g., reckless autonomous agents), and massive energy consumption of large models, arguing that AI must become as safe and regulated as nuclear power and far more energy-efficient. [44-56][66-67]
– India has a unique opportunity to lead a “human-centered” AI revolution.
He references the Prime Minister’s call for a “billion entrepreneurs,” India’s past successes (Green Revolution, telecom expansion), and the nation’s abundant talent as the foundation for building the next generation of AI that empowers people rather than merely automates tasks. [40-42][68-73]
Overall purpose / goal
The speaker’s aim is to motivate the audience-particularly Indian technologists, entrepreneurs, and policymakers-to harness AI’s unprecedented productivity, to responsibly close the gap between AI capabilities and business needs, to confront the technology’s safety and sustainability challenges, and ultimately to position India as a global leader in a purposeful, human-centric AI transformation.
Overall tone
– Opening (0:00-2:00): Warm, appreciative, and celebratory, thanking the previous speaker and praising Vishal Sikka’s credentials.
– Middle (2:00-7:00): Energetic and optimistic when describing productivity breakthroughs, then becomes more analytical and cautionary as he outlines the need for trustworthy systems and the “jagged frontier” of AI effectiveness.
– Later (7:00-10:42): Shifts to a serious, almost urgent tone when discussing AI’s limitations-hallucinations, safety, and energy concerns-while still maintaining a visionary optimism about overcoming these hurdles.
– Closing: Returns to an inspiring, hopeful tone, envisioning a “human revolution” powered by good AI and a billion Indian entrepreneurs, ending on a note of excitement and fun.
The tone evolves from gratitude to excitement, through sober analysis of risks, and culminates in a hopeful call to action.
Speakers
– Vishal Sikka – Founder and CEO of VNI; former CEO of Infosys; computer scientist and AI thought leader. [S2]
– Speaker 1 – Event moderator/host (role not specified). [S3]
Additional speakers:
– None identified.
The session opened with a formal note of appreciation. Speaker 1 thanked Sir Hasabis for his “profound and illuminating address” and expressed sincere gratitude before introducing the next speaker, Vishal Sikka, founder and CEO of VNI, a former Infosys chief who led “one of the most ambitious transformations in Indian IT history” and is described as “a computer scientist by training, a philosopher by temperament” and a leading thinker at the AI-enterprise intersection [1-10].
Sikka began by echoing the courteous tone, thanking the audience twice and calling the event “wonderful” [11-13]. He then outlined that his remarks would be organised around three main observations drawn from his long experience in artificial intelligence [14].
First observation – productivity gains
He recounted a Stanford classmate who rebuilt a nine-month, 15-engineer service in 14 days using a generative-coding tool, achieving >250× productivity [15-20]. A second illustration involved a home-goods distributor who, with VNI’s product, reduced a decision that would normally require a year of analysis to a few days, providing instant, multilingual knowledge synthesis [20-22]. These anecdotes underscore the tangible, measurable impact of AI when wielded by knowledgeable practitioners [15-22].
Second observation – bridging the LLM-enterprise gap
Sikka argued that effective AI use requires not only technical know-how but also awareness of large-language models (LLMs) limitations, and that enterprises need “trusted, verifiable, reliable systems” that sit above the LLMs to deliver correct business value [27-32]. VNI is developing such a layer that abstracts the underlying models while ensuring correctness and reliability [31-32]. He linked this bridging effort to broader transformation potential: by closing the gap, “legacy systems” and “enormous complexities inside enterprises” can be removed, allowing industries to be reshaped and giving end users wings and amplifying them [33-38].
Third observation – mastering AI’s limitations
Sikka identified “hallucinations” as a primary obstacle, noted the lack of physical-world understanding, and warned of safety risks such as reckless autonomous swarms [44-56]. He likened the need for AI safety to the decades-long regulatory regime governing nuclear power, arguing that comparable rigor is essential for responsible AI deployment [55-56].
Energy consumption was presented as another critical constraint. Describing a 720 MW data centre on California’s Highway 101, he compared the power draw of generative-AI inference to a 100-watt light bulb and highlighted that the human brain operates on merely 15-20 W, implying that “many zeros still need to be removed” from today’s models [62-66]. To illustrate the inefficiency, he told the “32 000-step / two-burger” story, showing how far AI’s energy use is from human efficiency [69].
Sikka then invoked a quotation from the Sanskrit scripture Bhaj Govindam – “Samprapte sanihite kale…” – to stress that “knowledge without wisdom does not save us” [57-58]. He framed this cultural insight as a reminder that practical wisdom must accompany technical mastery.
Turning to the Indian context, Sikka invoked the Prime Minister’s call for “a billion entrepreneurs” who can harness AI to create value, stressing that India possesses abundant talent and an entrepreneurial spirit suited to this challenge [40-43]. He recalled the Green Revolution, which turned India from a food-deficit nation into a major exporter within a generation, and cited the nation-wide connectivity achieved by Jio and Airtel, now providing “billion-plus Indians” with data and internet access [70-73]. These precedents reinforce his belief that AI can be the next leap-frog for the country.
In closing, Sikka painted a vision of a “human revolution powered by AI, by good AI, by purposeful AI,” where a billion entrepreneurs are not merely earning a livelihood but “making a life” for themselves and others [74-75]. He reaffirmed the excitement of this endeavour, thanked the audience, and ended on an upbeat note, inviting participants to join the journey toward a responsible, empowering AI future [76].
Thank you so much, Sir Hasabis, for your very profound and illuminating address. We really thank you. Sincere gratitude to your address. Ladies and gentlemen, and now I would like to invite Mr. Vishal Sikka. He’s the founder and CEO of VNI. As CEO of Infosys, Vishal Sikka has led one of the most ambitious transformations in Indian IT history. Before leaving to build VNI, a company focused on human -centered artificial intelligence. He is a computer scientist by training, a philosopher by temperament. He is one of the most original thinkers of the intersection of AI and enterprise. Please welcome the founder and CEO of VNI, Mr. Vishal Sikka.
Thank you so much. Thank you so much. Wow, wonderful introduction and what an amazing event. I want to share three points from a long time that I have spent in AI. My first point is that what we see today in the world of AI, people who know what they are doing with AI are astonishingly effective with AI. Recently, a friend of mine, he and I were students together at Stanford. He has a large service that he runs, open public service. That service was built by 15 people, very world -class engineers, over nine months. Recently, he rebuilt that service entirely by himself in 14 days using one of the generative AI coding tools. So if you are counting, that is about more than 250 times improvement in productivity.
now he’s a genius and not everyone gets a 250 times productivity gain but you will see that people who understand how to use ai are astonishingly effective with it and i had a similar experience recently with a customer of mine who is a distributor of home goods and one of their main suppliers shut down their factories in one of their countries and they did analysis using our product all kinds of simulation scenarios and over a few days they reached the decision that they need to exit that country entirely i asked him you know such a monumental decision to get out of an entire country how long would that have taken you before and he said easily it would take you a year to get out of that country and i said you know what i mean by that it would have taken a year and it would have involved heavy duty consultants and things like that So, we now have instant access to knowledge in any language, a condensation of things that we can present in any way.
It is an incredible power. And yes, it is deeply disruptive to the ways that we have worked before, the way that we have done things in the past. But at the same time, and even more importantly, we can do unprecedented things with this, things that we could never do before. So, this astonishing effectiveness, and Yoshua referred to this as a jagged frontier, it is not uniform. Not everyone sees this. So, that’s my second point. Being effective with AI requires not only a knowledge of AI itself, but understanding its limitations and how to overcome those limitations. There is a huge gap between LLMs. and the business users inside enterprises especially and how to bring value to those users.
And overcoming that gap is where a lot of value -creating opportunity is. Bridging that gap requires delivering correct systems, trusted, verifiable, reliable systems that deliver value to people. My own company works in this area, a layer that sits above the language models and delivers value to business users while ensuring correctness and things like that. When we overcome that gap, we can deliver massive value. Mukesh Bhai talked about it just now. We can transform every existing system, legacy systems, enormous complexities inside enterprises can be removed. Industries can be transformed. We can give end users wings. We can end users. We can amplify them. to deliver things that were not possible to do before, that or in the best case, it required professionals to do this.
Doing that also requires not just overcoming the limitations of AI, but also imagination to see what is not there, to see what is possible. India has all of this in great abundance. The Honorable Prime Minister has called for a billion entrepreneurs, people who can overcome these and deliver value using AI. And I think this is exactly what the world needs and what India has the potential for. My third point is that we not only have to master today’s AI, but we have to leapfrog it. AI today has enormous limitations. I have worked in AI for the last 38 years. You know… One of our scriptures is Bhaj Govindam. It was written by the Shankaracharya. And it has a beautiful line.
Samprapte sanihite kale, nahi nahi rakshati dukhrin karane. What it means is that when you are faced with a life or death situation, the knowledge of a book does not help you. Knowledge without wisdom does not save us. That wisdom comes from living, from doing, from being in the world. AI today has plenty of limitations. Yoshua mentioned hallucinations. That’s one of the main issues blocking the use of AI in enterprises. But beyond hallucinations, understanding the world, understanding physical activities, the physical movement, this is one of the next frontiers. Safety. safety of AI and the Honorable Prime Minister talked about this today is an existential issue swarms of agents can be made to do completely reckless things and we don’t yet have ways to understand or deal with this and Yoshua Demis also talked about this we have to solve this issue we have to deliver AI that is safe we have done this with nuclear power for the last 80 plus years we can and we must do this with AI energy is another one of these issues where I live in California on highway 101 just north of San Tomas I drive by there every time I go to see my dad there is a massive data center that is coming up it’s 720 megawatts and this idea that I write a prompt and these gazillion genes GPUs blast into existence to produce a response and then I make a tiny change to that prompt and then I do that again.
It just seems like a completely absurd idea, especially to someone who has been around AI for such a long time. I have, thanks to the minister, Ministry of IT, Honorable Ashwini Vasanthaji, I have a colleague who has been accompanying me throughout this conference and he told me that yesterday he walked 32 ,000 steps. And I asked him, what did you eat? And he said, I ate two burgers. Shubham, if you’re here. Two burgers. You know, we normally eat 2 ,000 calories in a day. That’s about a 100 -watt light bulb, like less than one of these light bulbs. And out of that, our brain, our nervous system is maybe 15, 20 watts. That’s like when your laptop is in sleep mode, it takes more power than that.
So, there are… many zeros still to be removed from these models, and the models themselves have to be removed. So, I think that there is a tremendous opportunity here. India is a country of the human potential. We have plenty of times before delivered the ability to, you know, billion plus Indians. Mukesh Bhai talked about Jio, and Sunil talked about Airtel, and how now we have billion plus Indians who have data and connectivity. When I was young, one of my earliest childhood memories is of worry in my parents’ faces around food. They used to tell stories of how there was a shortage of food, and then the green revolution happened, and India is now one of the largest exporters of food in one generation.
So, I think that when you look at the time of intelligence, it is not only an opportunity to learn about this technology, to learn to master this technology, to understand its limitations, but to leapfrog that and to build the next generation of it. And as this summit so vividly demonstrates, we can be on our way to a human revolution powered by AI, by good AI, by purposeful AI, where every one of us, a billion entrepreneurs, is not just making a living but is making a life, not some artificial life or some artificial general life, but our own life and the life of others. And that would be so much fun. Fun to do. Thank you so much.
“Vishal Sikka is the founder and CEO of VNI and a former Infosys chief who led one of the most ambitious transformations in Indian IT history”
The knowledge base records that Vishal Sikka is the founder and CEO of VNI, former CEO of Infosys, and that he led one of the most ambitious transformations in Indian IT history [S1] and [S2].
“Sikka is described as “a computer scientist by training, a philosopher by temperament””
The moderator introduced Sikka with exactly those descriptors in the source material [S2].
“Sikka delivered a keynote address on artificial intelligence’s transformative potential and challenges”
The knowledge base notes that Sikka gave a keynote on AI’s transformative potential and challenges, matching the report’s description of his talk [S2].
The only clear point of agreement between the two speakers is the mutual expression of thanks at the outset of the session; substantive AI‑related arguments are presented solely by Vishal Sikka with no corresponding statements from Speaker 1.
Low consensus on policy or technical issues; the limited agreement is limited to procedural courtesy, suggesting that the discussion did not produce substantive shared positions on AI, safety, productivity, or broader development themes.
The transcript shows no substantive disagreement between the speakers. Speaker 1’s remarks are limited to gratitude and the formal introduction of Vishal Sikka ([1-4]), while Sikka’s extensive remarks focus on AI productivity gains, the need for trustworthy layers, safety and sustainability concerns ([14-24][27-33][54-57]). Their statements are complementary rather than contradictory, indicating alignment on the relevance and potential of AI for India.
Minimal – the interaction is largely harmonious, implying consensus on the importance of AI and its potential impact, which suggests a unified stance for the topics under discussion.
Vishal Sikka’s remarks steered the discussion from a surface‑level celebration of AI capabilities to a layered exploration of productivity, inequality, trust, safety, sustainability, and national ambition. Each pivotal comment introduced a fresh dimension—whether technical (250× productivity), philosophical (knowledge vs. wisdom), regulatory (AI safety akin to nuclear power), or societal (leap‑frogging like the Green Revolution). These insights acted as turning points that redirected attention, deepened analysis, and broadened the conversation’s scope, ultimately shaping the dialogue into a balanced narrative that combined optimism with caution and positioned AI as a catalyst for a large‑scale human transformation.
Disclaimer: This is not an official session record. DiploAI generates these resources from audiovisual recordings, and they are presented as-is, including potential errors. Due to logistical challenges, such as discrepancies in audio/video or transcripts, names may be misspelled. We strive for accuracy to the best of our ability.
Related event
