Keynote-Vishal Sikka
19 Feb 2026 12:00h - 12:15h
Keynote-Vishal Sikka
Session at a glance
Summary
Vishal Sikka, founder and CEO of VNI and former CEO of Infosys, delivered a keynote address on artificial intelligence’s transformative potential and challenges at what appears to be an AI summit in India. Sikka presented three main points based on his 38 years of experience in AI development and implementation. His first point emphasized that people who truly understand AI can achieve astonishing effectiveness, citing examples of a 250-times productivity improvement when his Stanford colleague rebuilt a service in 14 days using generative AI tools that originally took 15 engineers nine months to create. He also described how a customer used AI to make a critical business decision about exiting an entire country in just days rather than the typical year-long process involving expensive consultants.
Sikka’s second point addressed the significant gap between AI capabilities and practical business applications, noting that bridging this gap requires delivering correct, trusted, and verifiable systems while understanding AI’s limitations. He emphasized that India has the potential to produce a billion entrepreneurs who can overcome these challenges and deliver AI value, referencing the Prime Minister’s vision. His third point focused on the need to not only master current AI but to leapfrog its limitations, including issues like hallucinations, safety concerns, and enormous energy consumption. Drawing from Indian scripture and comparing human brain efficiency to current AI models, Sikka argued that there are tremendous opportunities to improve AI technology. He concluded by envisioning a “human revolution powered by AI” where India could lead the development of purposeful AI that enhances human potential rather than replacing it.
Keypoints
Major Discussion Points:
– Astonishing effectiveness of AI for knowledgeable users: Sikka emphasizes that people who understand AI can achieve extraordinary productivity gains, citing examples of 250x improvement in development speed and rapid business decision-making that previously would have taken a year with consultants.
– The gap between AI capabilities and business implementation: There’s a significant challenge in bridging the divide between large language models and actual business users in enterprises, requiring trusted, verifiable, and reliable systems that can deliver real value while overcoming AI’s limitations.
– Current limitations of AI technology: Discussion of major obstacles including hallucinations, lack of understanding of physical world activities, safety concerns with AI agents, and the enormous energy consumption required by current AI systems.
– India’s potential to leapfrog AI development: Sikka argues that India has the human potential, imagination, and historical precedent (citing examples like the Green Revolution and telecom transformation) to not just adopt current AI but to build the next generation of AI technology.
– Vision for human-centered AI revolution: The ultimate goal of creating “good AI” and “purposeful AI” that enables a billion entrepreneurs to not just make a living but create meaningful lives, moving beyond artificial general intelligence to enhance human potential.
Overall Purpose:
The discussion aims to present a strategic vision for how India can approach AI development and implementation, moving beyond simply adopting existing AI technologies to becoming a leader in creating the next generation of human-centered artificial intelligence.
Overall Tone:
The tone is consistently optimistic and inspirational throughout, with Sikka maintaining an encouraging perspective even when discussing AI’s limitations and challenges. He balances realism about current obstacles with confidence in India’s potential, using personal anecdotes and cultural references to create an uplifting narrative about technological transformation and human empowerment.
Speakers
– Moderator: Role/Title: Event moderator; Area of expertise: Not mentioned
– Vishal Sikka: Role/Title: Founder and CEO of VNI, former CEO of Infosys; Area of expertise: Computer scientist, artificial intelligence, enterprise AI, human-centered artificial intelligence (described as “a philosopher by temperament” and “one of the most original thinkers of the intersection of AI and enterprise”)
Additional speakers:
– Sir Hassabis: Role/Title: Not specified; Area of expertise: Not mentioned (referenced as having given a previous address)
– Mukesh Bhai: Role/Title: Not specified; Area of expertise: Not mentioned (referenced in relation to Jio)
– Sunil: Role/Title: Not specified; Area of expertise: Not mentioned (referenced in relation to Airtel)
– Honorable Prime Minister: Role/Title: Prime Minister; Area of expertise: Not mentioned (referenced for calling for “a billion entrepreneurs” and discussing AI safety)
– Honorable Ashwini Vasanthaji: Role/Title: Minister, Ministry of IT; Area of expertise: Information Technology
– Yoshua: Role/Title: Not specified; Area of expertise: AI (referenced discussing “jagged frontier” and hallucinations)
– Demis: Role/Title: Not specified; Area of expertise: AI (referenced discussing AI safety)
Full session report
Vishal Sikka, founder and CEO of VNI and former CEO of Infosys, delivered a keynote address on artificial intelligence’s transformative potential and challenges. The moderator introduced Sikka as “a computer scientist by training, a philosopher by temperament” and “one of the most original thinkers at the intersection of AI and enterprise.” Drawing from his 38 years of experience in AI, Sikka presented a three-point framework addressing AI’s extraordinary capabilities, implementation challenges, and future limitations.
The Astonishing Effectiveness of AI for Knowledgeable Users
Sikka’s first point centered on the remarkable productivity gains achievable by those who understand how to leverage AI effectively. He provided compelling evidence through personal anecdotes, most notably involving a Stanford colleague who had originally built a large public service platform using 15 world-class engineers over nine months. Recently, this same individual rebuilt the entire service single-handedly in just 14 days using generative AI coding tools—a 250-times improvement in productivity.
Sikka reinforced this with another example involving a home goods distributor facing a crisis when a main supplier shut down factories. Using AI-powered analysis through Sikka’s company’s product, the distributor made the decision to exit that entire market within days—a process that would traditionally have required at least a year and expensive consultants.
These examples demonstrated what Sikka characterized as “instant access to knowledge in any language” and “incredible power” that enables “unprecedented things.” However, he noted this effectiveness follows what Yoshua (presumably Bengio) referred to as a “jagged frontier”—the benefits are not uniformly distributed across all users and applications.
Bridging the Gap Between AI Capabilities and Business Implementation
Sikka’s second point addressed the substantial gap between the raw capabilities of large language models and the practical needs of business users. He argued that being effective with AI requires understanding not only the technology but crucially its limitations and how to overcome them.
This gap represents “where a lot of value-creating opportunity is,” requiring systems that are correct, trusted, verifiable, and reliable while providing tangible value. Sikka’s company operates in this space, developing “a layer that sits above the language models and delivers value to business users while ensuring correctness.”
When organizations successfully bridge this gap, they can deliver massive transformation. Sikka referenced comments from other speakers, including “Mukesh Bhai” and Sunil’s remarks about Airtel, regarding transforming existing systems and removing complexities from legacy enterprise systems. The goal is to “give end users wings” and amplify their capabilities.
Sikka positioned India as uniquely equipped for this challenge, referencing the Prime Minister’s vision of creating “a billion entrepreneurs” and arguing this represents exactly what the world needs and what India can provide.
The Imperative to Leapfrog Current AI Limitations
Sikka’s third point focused on moving beyond today’s AI limitations to build next-generation systems. Drawing from ancient Indian wisdom, he quoted from Shankaracharya’s Bhaj Govindam: “Samprapte sanihite kale, nahi nahi rakshati dukhrin karane,” which he translated as “when you are faced with a life or death situation, the knowledge of a book does not help you.” He interpreted this as “knowledge without wisdom does not save us,” with wisdom coming from “living, from doing, from being in the world.”
This led Sikka to identify critical limitations in current AI systems. Beyond hallucinations—”one of the main issues blocking the use of AI in enterprises”—he highlighted AI’s poor understanding of the physical world and safety concerns, noting that “swarms of agents can be made to do completely reckless things.” He referenced previous speakers Demis (presumably Hassabis) and Yoshua regarding these safety challenges.
Perhaps most provocatively, Sikka addressed AI’s enormous energy consumption, calling it “completely absurd.” He described driving past a massive 720-megawatt data center being constructed, contrasting this with human cognitive efficiency. Humans consume about 2,000 calories daily (equivalent to a 100-watt light bulb), with the brain operating on merely 15-20 watts—less than a laptop in sleep mode. This highlighted fundamental inefficiencies suggesting “many zeros still to be removed from these models, and the models themselves have to be removed.”
India’s Unique Position for AI Leadership
Throughout his address, Sikka positioned India as uniquely positioned to lead AI development. He shared a personal childhood memory of seeing worry in his parents’ faces about food shortages, then witnessing India’s transformation through the Green Revolution into a major food exporter within a generation. He drew parallels to recent transformations, referencing how Indian telecommunications leaders connected over a billion Indians with data and connectivity.
Sikka envisioned a “human revolution powered by AI”—specifically “good AI” and “purposeful AI”—enabling a billion entrepreneurs to not merely “make a living” but to “make a life.” His conclusion emphasized creating “not some artificial life or some artificial general life, but our own life and the life of others.”
The address demonstrated Sikka’s ability to weave together technical examples, philosophical insights, and personal anecdotes into a coherent vision for human-centered AI development. His balanced approach acknowledged both extraordinary potential and significant limitations, providing a framework for AI that amplifies rather than replaces human capabilities. He concluded enthusiastically: “And that would be so much fun. Fun to do. Thank you so much.”
Session transcript
Thank you so much, Sir Hassabis, 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.
Moderator
Speech speed
128 words per minute
Speech length
110 words
Speech time
51 seconds
Opening and transition
Explanation
The moderator thanks the previous speaker and formally introduces Vishal Sikka, highlighting his leadership at Infosys and setting the stage for his remarks.
Evidence
“Ladies and gentlemen, and now I would like to invite Mr. Vishal Sikka.” [44]. “Please welcome the founder and CEO of VNI, Mr. Vishal Sikka.” [45]. “As CEO of Infosys, Vishal Sikka has led one of the most ambitious transformations in Indian IT history.” [46]. “Sincere gratitude to your address.” [47].
Major discussion point
Opening and transition
Topics
Information and communication technologies for development | The enabling environment for digital development
Vishal Sikka
Speech speed
134 words per minute
Speech length
1329 words
Speech time
592 seconds
250‑times productivity gain with generative AI
Explanation
Sikka reports that rebuilding a service with generative AI tools in 14 days delivered more than a 250‑fold increase in productivity compared with the traditional nine‑month effort by a large engineering team.
Evidence
“So if you are counting, that is about more than 250 times improvement in productivity.” [1]. “Recently, he rebuilt that service entirely by himself in 14 days using one of the generative AI coding tools.” [2]. “That service was built by 15 people, very world‑class engineers, over nine months.” [4].
Major discussion point
AI productivity and transformative impact
Topics
Artificial intelligence | The digital economy
Year‑long decision reduced to days
Explanation
Using AI‑driven simulations, a home‑goods distributor was able to decide to exit an entire country in a few days, a process that would have taken roughly a year with traditional consulting.
Evidence
“…a distributor of home goods … 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 … he said easily it would take you a year to get out of that country…” [3].
Major discussion point
AI productivity and transformative impact
Topics
Artificial intelligence | Social and economic development
Trusted, verifiable systems above LLMs
Explanation
Sikka emphasizes that bridging the gap between large language models and enterprise value requires a layer of trusted, verifiable systems that ensure correctness for business users.
Evidence
“Bridging that gap requires delivering correct systems, trusted, verifiable, reliable systems that deliver value to people.” [23]. “There is a huge gap between LLMs.” [24]. “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.” [25].
Major discussion point
Bridging AI limitations to deliver enterprise value
Topics
Artificial intelligence | The enabling environment for digital development
India’s entrepreneurial pool for AI value
Explanation
Sikka points to India’s vast human potential and the government’s call for a billion AI‑savvy entrepreneurs as a catalyst for delivering AI‑driven value at scale.
Evidence
“The Honorable Prime Minister has called for a billion entrepreneurs, people who can overcome these and deliver value using AI.” [27]. “India is a country of the human potential.” [29]. “India has all of this in great abundance.” [31].
Major discussion point
Bridging AI limitations to deliver enterprise value
Topics
Artificial intelligence | Capacity development | The enabling environment for digital development
Safety and hallucinations as barriers
Explanation
Sikka flags AI safety, including hallucinations, as existential risks for enterprises and calls for safety standards comparable to those used in nuclear power.
Evidence
“Safety.” [34]. “Safety of AI … is an existential 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” [35]. “Yoshua mentioned hallucinations.” [37].
Major discussion point
Future challenges, safety, and the need to leapfrog AI
Topics
Artificial intelligence | Building confidence and security in the use of ICTs | Human rights and the ethical dimensions of the information society
Energy consumption of large models
Explanation
Sikka illustrates the massive energy demand of current AI models by referencing a 720‑megawatt data centre, underscoring the need for more efficient AI architectures.
Evidence
“…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…” [35].
Major discussion point
Future challenges, safety, and the need to leapfrog AI
Topics
Environmental impacts | Artificial intelligence
Wisdom over knowledge, need to leapfrog
Explanation
Sikka argues that knowledge without wisdom is insufficient and urges the community to not only master today’s AI but to leapfrog it and build the next generation.
Evidence
“Knowledge without wisdom does not save us.” [42]. “My third point is that we not only have to master today’s AI, but we have to leapfrog it.” [22]. “…but to leapfrog that and to build the next generation of it.” [28].
Major discussion point
Future challenges, safety, and the need to leapfrog AI
Topics
Artificial intelligence | Human rights and the ethical dimensions of the information society
Agreements
Agreement points
AI has transformative potential but requires addressing significant limitations
Speakers
– Vishal Sikka
Arguments
Current AI Effectiveness and Productivity Gains – People who understand AI are achieving extraordinary productivity improvements, with examples of 250x productivity gains in software development and rapid business decision-making that previously took a year now completed in days
AI Limitations and Future Challenges – Current AI has major limitations including hallucinations, lack of understanding of physical world activities, and safety concerns with AI agent swarms
Summary
Sikka acknowledges both the extraordinary potential of AI for productivity gains while simultaneously recognizing its current limitations that must be overcome
Topics
Artificial intelligence | Building confidence and security in the use of ICTs
India has unique potential for AI leadership and development
Speakers
– Vishal Sikka
Arguments
India’s Potential in AI Development and Leadership – India has the knowledge, imagination, and human potential to not only master current AI but to leapfrog it and build next-generation AI systems
India can lead a human revolution powered by purposeful AI, creating a billion entrepreneurs who make meaningful lives rather than just artificial intelligence
Summary
Sikka consistently argues that India is positioned to lead in AI development and create a human-centered AI revolution
Topics
Artificial intelligence | Social and economic development | The enabling environment for digital development
Similar viewpoints
Both arguments emphasize the critical need for reliable, trustworthy AI systems that can deliver practical value to business users, highlighting the gap between current AI capabilities and enterprise needs
Speakers
– Vishal Sikka
Arguments
Bridging the Gap Between AI Technology and Business Users – There is a significant gap between large language models and enterprise business users that needs to be bridged through correct, trusted, and verifiable systems
Overcoming AI limitations requires delivering reliable systems that provide real value to business users while ensuring correctness
Topics
Artificial intelligence | The digital economy | Building confidence and security in the use of ICTs
Both arguments identify fundamental inefficiencies and limitations in current AI systems that need to be addressed for sustainable and safe AI development
Speakers
– Vishal Sikka
Arguments
AI Limitations and Future Challenges – Current AI has major limitations including hallucinations, lack of understanding of physical world activities, and safety concerns with AI agent swarms
Energy consumption of current AI systems is extremely inefficient compared to human brain processing, requiring massive data centers and computational resources
Topics
Artificial intelligence | Environmental impacts | Building confidence and security in the use of ICTs
Unexpected consensus
Energy efficiency criticism of current AI systems
Speakers
– Vishal Sikka
Arguments
Energy consumption of current AI systems is extremely inefficient compared to human brain processing, requiring massive data centers and computational resources
Explanation
It’s unexpected for an AI advocate and industry leader to be so critical of current AI energy consumption, comparing it unfavorably to human brain efficiency and calling the current approach ‘completely absurd’
Topics
Artificial intelligence | Environmental impacts
Overall assessment
Summary
The transcript contains only one speaker (Vishal Sikka) presenting arguments, so traditional consensus analysis between multiple speakers is not applicable. However, Sikka demonstrates internal consistency in his arguments, balancing optimism about AI’s potential with realistic acknowledgment of its limitations. His arguments show agreement on the need for human-centered, reliable AI systems and India’s potential for AI leadership.
Consensus level
Since only one substantive speaker is present, there is inherent consistency in the viewpoints presented. Sikka’s arguments form a coherent framework that acknowledges both AI’s transformative potential and its current limitations, suggesting a balanced approach to AI development and implementation.
Differences
Different viewpoints
Unexpected differences
Overall assessment
Summary
No disagreements identified as the transcript contains only one substantive speaker (Vishal Sikka) presenting his perspective on AI effectiveness, limitations, and India’s potential in AI development. The moderator only provides an introduction without presenting any arguments or viewpoints.
Disagreement level
No disagreement present. This is a single-speaker presentation rather than a debate or discussion with multiple viewpoints. Sikka presents a cohesive narrative about AI’s current state, challenges, and future potential without any opposing voices or contrasting perspectives being presented.
Partial agreements
Partial agreements
Similar viewpoints
Both arguments emphasize the critical need for reliable, trustworthy AI systems that can deliver practical value to business users, highlighting the gap between current AI capabilities and enterprise needs
Speakers
– Vishal Sikka
Arguments
Bridging the Gap Between AI Technology and Business Users – There is a significant gap between large language models and enterprise business users that needs to be bridged through correct, trusted, and verifiable systems
Overcoming AI limitations requires delivering reliable systems that provide real value to business users while ensuring correctness
Topics
Artificial intelligence | The digital economy | Building confidence and security in the use of ICTs
Both arguments identify fundamental inefficiencies and limitations in current AI systems that need to be addressed for sustainable and safe AI development
Speakers
– Vishal Sikka
Arguments
AI Limitations and Future Challenges – Current AI has major limitations including hallucinations, lack of understanding of physical world activities, and safety concerns with AI agent swarms
Energy consumption of current AI systems is extremely inefficient compared to human brain processing, requiring massive data centers and computational resources
Topics
Artificial intelligence | Environmental impacts | Building confidence and security in the use of ICTs
Takeaways
Key takeaways
AI mastery creates extraordinary productivity gains – experts are achieving 250x improvements in development speed and reducing year-long business decisions to days
There exists a critical gap between AI technology capabilities and practical business implementation that must be bridged through reliable, trusted systems
Current AI has significant limitations including hallucinations, poor physical world understanding, safety risks with agent swarms, and extremely inefficient energy consumption compared to human cognition
India has exceptional potential to not just adopt current AI but to leapfrog existing technology and lead next-generation AI development
The ultimate goal should be a ‘human revolution powered by AI’ that creates meaningful lives and empowers a billion entrepreneurs, rather than pursuing artificial general intelligence for its own sake
Resolutions and action items
India should focus on creating a billion entrepreneurs who can master AI and deliver value using the technology
Need to develop AI systems that are safe, reliable, and energy-efficient to overcome current limitations
Must build bridging layers above language models to deliver practical value to business users while ensuring correctness
Unresolved issues
How to specifically solve AI hallucination problems that block enterprise adoption
Methods for ensuring AI safety and controlling potentially reckless AI agent swarms
Technical approaches for dramatically reducing AI energy consumption to sustainable levels
Specific strategies for how India can leapfrog current AI technology to build next-generation systems
Concrete implementation details for bridging the gap between LLMs and business users
Suggested compromises
None identified
Thought provoking comments
Recently, a friend of mine… rebuilt that service entirely by himself in 14 days using one of the generative AI coding tools… that is about more than 250 times improvement in productivity.
Speaker
Vishal Sikka
Reason
This concrete example provides a stark, quantifiable demonstration of AI’s transformative potential. The 250x productivity gain is not theoretical but based on real experience, making it particularly compelling and credible.
Impact
This comment establishes the foundation for the entire discussion by providing tangible evidence of AI’s disruptive power. It shifts the conversation from abstract possibilities to concrete realities, setting up the framework for discussing both opportunities and challenges.
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.
Speaker
Vishal Sikka
Reason
This insight reveals the nuanced reality behind AI adoption – that raw capability doesn’t automatically translate to business value. It highlights the critical implementation gap that many overlook when discussing AI’s potential.
Impact
This comment introduces complexity to what could have been a purely optimistic narrative about AI. It acknowledges the ‘jagged frontier’ concept and pivots the discussion toward practical challenges, making the conversation more balanced and realistic.
Samprapte sanihite kale, nahi nahi rakshati dukhrin karane… Knowledge without wisdom does not save us. That wisdom comes from living, from doing, from being in the world.
Speaker
Vishal Sikka
Reason
This philosophical insight, drawing from ancient Sanskrit scripture, provides a profound critique of current AI limitations. It distinguishes between mere information processing and true understanding, highlighting AI’s fundamental gap in experiential wisdom.
Impact
This comment elevates the discussion from technical considerations to philosophical depths, introducing the concept that AI’s current limitations are not just technical but fundamental to the nature of knowledge versus wisdom. It adds a uniquely Indian philosophical perspective to the global AI discourse.
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… It just seems like a completely absurd idea… our brain, our nervous system is maybe 15, 20 watts.
Speaker
Vishal Sikka
Reason
This comparison between AI’s energy consumption and human brain efficiency is striking and thought-provoking. The contrast between 720 megawatts for a data center versus 15-20 watts for human cognition highlights the inefficiency of current AI systems.
Impact
This observation introduces a critical sustainability perspective to the AI discussion, suggesting that current approaches are fundamentally flawed. It challenges the audience to think about AI development not just in terms of capability but also efficiency and environmental impact.
When I was young, one of my earliest childhood memories is of worry in my parents’ faces around food… and then the green revolution happened, and India is now one of the largest exporters of food in one generation.
Speaker
Vishal Sikka
Reason
This personal anecdote connects historical transformation to current AI potential, providing both emotional resonance and historical precedent for dramatic societal change. It demonstrates India’s capacity for leapfrog development.
Impact
This comment provides inspirational context and historical grounding for the discussion, suggesting that India has successfully navigated transformative technologies before. It shifts the tone toward optimism while acknowledging the magnitude of change possible within a single generation.
Overall assessment
Vishal Sikka’s speech masterfully weaves together concrete examples, philosophical insights, and historical perspective to create a nuanced narrative about AI’s potential and limitations. His comments shaped the discussion by moving it through three distinct phases: establishing AI’s transformative power through quantifiable examples, introducing complexity by acknowledging limitations and implementation challenges, and finally providing both philosophical depth and historical context for optimistic yet realistic expectations. The integration of ancient Sanskrit wisdom with cutting-edge technology discussion demonstrates a uniquely Indian approach to AI discourse, while the personal anecdotes and specific examples ground abstract concepts in relatable human experience. This creates a comprehensive framework that acknowledges both the revolutionary potential and the significant challenges of AI adoption.
Follow-up questions
How can we bridge the gap between LLMs and business users in enterprises to deliver reliable, verifiable systems?
Speaker
Vishal Sikka
Explanation
This addresses a critical barrier to AI adoption in enterprises, where there’s a significant disconnect between AI capabilities and practical business applications that require correctness and trust
How can we solve the hallucination problem that is blocking AI use in enterprises?
Speaker
Vishal Sikka
Explanation
Hallucinations are identified as one of the main technical issues preventing widespread enterprise adoption of AI systems, requiring urgent research and solutions
How can we develop AI systems that better understand the physical world and physical activities?
Speaker
Vishal Sikka
Explanation
Current AI limitations in understanding physical movement and real-world interactions represent a key frontier for advancing AI capabilities beyond text and language processing
How can we ensure AI safety, particularly with swarms of agents that can be made to do reckless things?
Speaker
Vishal Sikka
Explanation
AI safety is described as an existential issue, with the potential for AI agent swarms to cause harm without proper safeguards and understanding
How can we dramatically reduce the energy consumption of AI models and systems?
Speaker
Vishal Sikka
Explanation
The massive energy requirements of current AI systems (exemplified by 720-megawatt data centers) are unsustainable compared to human brain efficiency, requiring research into more efficient architectures
How can we develop the next generation of AI that overcomes current limitations and moves beyond existing models?
Speaker
Vishal Sikka
Explanation
This represents a call for fundamental research to leapfrog current AI technology rather than just incrementally improving existing systems
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
