Keynote-Vinod Khosla
19 Feb 2026 15:30h - 15:45h
Keynote-Vinod Khosla
Summary
The session opened by welcoming Silicon Valley investor Vinod Khosla to present his vision for rapid AI deployment in India [2][6-7]. Khosla emphasized that AI initiatives must first benefit the bottom half of the Indian population to achieve large-scale impact [12][13].
He highlighted AI-based personal tutors as an already-available solution, noting that millions of children in India currently use such tools [13][19-22]. According to Khosla, the CK-12 platform has reached about 400 million students worldwide, with 4 million Indian users and over 12 million regular users, all for free and aligned with national curricula [24-28][30-31]. He argued that AI tutors can assess a learner within minutes and tailor instruction through knowledge tracing, potentially outperforming human tutors [46-48][33].
Khosla then described AI-driven doctors that could provide 24/7 primary-care, disease management, mental-health, and nutrition coaching at negligible cost, surpassing services even in the United States [14-15][50-55][53]. He claimed that, aside from physical examinations, there is little a human doctor can do that current AI cannot, and that AI would triage cases to physicians when necessary [57-60]. To deliver these services, Khosla proposes creating a Section 8 nonprofit that integrates AI health, education, and agronomy platforms into the Aadhaar identity system, leveraging the same infrastructure that enabled UPI [60-63][64].
He also advocated for AI-level PhD agronomists accessible to every farmer via the same Aadhaar-linked model, allowing local, language-specific advice [16][68-69]. Khosla believes that scaling AI doctors could bring India’s doctor-patient ratio ahead of that of the United States within a few years, even without massive financial investment [71-74]. He envisions the AI systems initially supervised by physicians, akin to an intern, with oversight diminishing after two to three years as the technology matures [81-86].
Across education, health, and agriculture, Khosla asserts that these AI services can be deployed cheaply within one to two years, reaching the poorest segments of society and avoiding a massive opportunity loss [92-96]. The discussion concluded that the future of large-scale, low-cost AI applications in India is already present and ready for immediate implementation [92-95].
Keypoints
– AI-based personal tutoring for K-12 students – Khosla proposes deploying AI tutors that can assess a learner in minutes and fill knowledge gaps, arguing they are “far superior to human tutors” and already in use by millions of Indian students through platforms like CK-12 ([13][24-30][46-49]).
– AI-driven 24/7 primary healthcare – He outlines AI doctors that provide continuous primary-care, disease management, mental-health and nutrition coaching at almost no cost, capable of triaging to human physicians when needed, and claims they can dramatically improve India’s doctor-patient ratio, even surpassing U.S. levels ([14][50-58][71-76][80-86]).
– AI agronomy services for farmers – Khosla envisions every farmer having a “PhD-level agronomist” available via AI, integrated with a UPI-like system to deliver localized advice in all Indian languages ([16][68-70][89-90]).
– Integration with the Aadhaar identity platform and a nonprofit delivery model – He recommends building a Section-8 nonprofit to embed AI tutors, doctors, and agronomists into the Aadhaar ecosystem, leveraging the existing identity infrastructure that enabled UPI ([60-64][65-66]).
– Focus on the bottom half of the population and urgency of action – The speaker stresses that AI must benefit the “bottom half of the Indian population” to achieve massive impact, warning that failure to act would be a “massive opportunity loss” ([12][95-96]).
Overall purpose:
The discussion is a rallying call to launch large-scale, low-cost AI applications in education, health care, and agriculture in India within the next one-to-two years, using existing digital infrastructure (Aadhaar/UPI) and a nonprofit model to reach the country’s poorest citizens and unlock billions of lives of benefit.
Tone:
Khosla’s tone is consistently upbeat, confident, and urgent. He moves from an introductory promise of immediate action to detailed, optimistic descriptions of each AI service, and concludes with a persuasive, almost urgent appeal to seize the opportunity before it is lost. The tone remains enthusiastic throughout, with a slight shift from explanatory to rally-cry as the talk progresses.
Speakers
– Vinod Khosla – Founder of Khosla Ventures; Co-founder of Sun Microsystems; venture capitalist and investor focusing on AI, climate and healthcare innovations[S1].
– Speaker 1 – Event moderator/host who introduced the keynote speaker[S3][S5].
Additional speakers:
Speaker 1 opened the session by thanking Mr Chit Adani and introducing Vinod Khosla – founder of Khosla Ventures and co-founder of Sun Microsystems – as “one of Silicon Valley’s most visionary investors” who has long bet on AI, climate and health care [1-3]. He noted Khosla’s view that roughly 80 % of jobs may be automated, but that this should be framed as an opportunity rather than a threat [4]. Khosla then stated that AI-driven impact must first reach the lower-income half of India’s population or it will not generate large-scale change [12].
Education – AI-powered personal tutoring agents
Khosla described an AI-tutor platform that can assess a learner’s knowledge in ten to fifteen minutes and then fill gaps through “knowledge tracing” [48-49]. He claimed that these tutors could outperform human tutors and may enable better learning than private tutoring [33-35][46-48]. The service builds on CK-12, which already provides free AI-generated content to millions worldwide, with about four million Indian users and more than twelve million regular users [24-28]. It is compatible with the CBSE curriculum and with state standards in multiple languages (English, Hindi, Odia, Meghalaya, etc.) [30-32] and includes a teacher-professional-development curriculum [44-45]. Khosla explained that the existing Diksha platform is “mostly unusable” and that the AI-tutor will be delivered as a Diksha 3.0, AI-first experience [98-99]. He proposed embedding the tutoring service in the Aadhaar ecosystem, alongside the health and agronomy services, to achieve universal, low-cost delivery [110-111].
Health care – AI-driven 24/7 primary-care doctors
The proposed AI-driven primary-care system would handle diagnosis, test ordering, prescriptions, chronic-disease management, mental-health therapy, free physical therapy, and nutrition coaching [50-53][108-109]. Khosla said the technology rests on a five-year platform developed by a company and adapted to Indian languages through the Sarvam model [100-101]. At launch the dialogue will be physician-approved [102-103] and the AI will function like a fresh-graduate MBBS “intern” under doctor supervision for the first one to two years, after which it could operate more autonomously [104-105][81-86]. The system can triage cases to human physicians and trigger emergency-room referrals when needed [106-107]. Khosla projected that such a service could give India a doctor-patient ratio that surpasses that of the United States [71-74]. Delivery would be through a Section 8 nonprofit (a non-profit under Indian law) that builds the platform and then hands it over to the Aadhaar infrastructure, mirroring how Aadhaar enabled UPI [60-66][63-65][S1].
Agronomy – AI-enabled “PhD-level” agronomist for every farmer
Khosla envisioned an AI agronomy assistant that farmers could access via voice or image input in any Indic language, even if illiterate, providing personalized advice on crops, pests and soil [68-70][89-90][16-18]. The service would be linked to an Aadhaar/UPI-like infrastructure, allowing 24/7, low-cost advice [68-70][89-90]. Similar AI-agronomy platforms have been cited as scalable solutions for small-holder farmers [S38].
Implementation model
A Section 8 nonprofit would develop the three AI services, iterate them to accommodate regional disease patterns, linguistic diversity and agricultural conditions, and then transfer the platforms to the Aadhaar ecosystem for universal, negligible-marginal-cost delivery [63-66][65-66][S1]. Multiple iteration cycles are expected to fine-tune the solutions for local contexts [65-66].
Socio-economic impact and urgency
All three services target the lower-income half of India’s population, offering cheap, scalable solutions that could “leap-frog” richer nations in education, health and agriculture outcomes [92-94][71-74]. Khosla concluded with a rallying call: the future is already here, and failing to act now would constitute a massive opportunity loss [95-96][97].
In sum, Vinod Khosla’s presentation combined an optimistic assessment of AI’s transformative potential with concrete, existing tools (such as CK-12 and the proposed Diksha 3.0) and a clear policy lever-the Aadhaar identity platform-to deliver low-cost, large-scale services in education, health care and agronomy. By positioning AI as a means to empower the lower-income half of India’s society, he aligned with the introductory speaker’s view that AI-driven disruption should be embraced as an opportunity rather than a threat [4][12][95-96].
Thank you, Mr. Chit Adani, for sharing your insights with us and your vision, as well as for enriching this August gathering. Ladies and gentlemen, it’s my privilege to now welcome Mr. Vinod Khosla, founder of Khosla Ventures, co -founder of Sun Microsystems and one of Silicon Valley’s most visionary investors. Mr. Vinod Khosla has been making bold bets on AI, on climate and health care for decades. He has argued that AI will replace 80 % of the jobs and that this is cause for optimism rather than despair. How? Let’s listen to him. Please welcome the founder of Khosla Ventures, Mr. Vinod Khosla.
Good afternoon. I am going to talk to you about some applications of AI that should be done immediately. I’m not going to talk about business or technology or where it’s going. I’m going to talk to you about what can be done today. If I can get my slides on the screen. Okay, so I’m going to talk to you about what can be done today in the next year or two to reach a billion and a half people in this country with really impactful immediate benefits. And unless AI benefits the bottom half of the Indian population, we’re not going to see a huge amount of impact. So, the first thing I’m going to talk to you that’s possible today, and in fact millions of kids in India are using today, is AI -based personal tutors.
And I’m going to talk to you about 24 by 7 almost free doctors available to everybody through AI. This is not helping a doctor, this is building a doctor. And of course, every farmer should have AI -level PhD agronomists available to them in their local small plot. This is all possible, they don’t even need to know how to read and write, just speak and look and take pictures. So, let me start with AI tutors. There’s a lot of children in India. There’s a lot of children in India who don’t get much help. in their education. In fact, in rural India, teachers don’t often show up. So it’s very important that this kind of a service be available so every child has their destiny in their own hands.
Thank you. The screen wasn’t showing my slides. My wife has been running a non -profit, ck12 .org, that offers it now. These are worldwide usage. About 400 million students have already used this service of AI content, which is all free, and AI tutors. In India, 4 million students have benefited by using the AI tutor. More than 12 million have used it constantly. So it is already in widespread use. This is already CBSE compatible, the national education policy compatible. The curriculums available in English or Hindi or Odisha or Meghalaya in these state standards, there are plenty of studies to show that they can be very efficacious. Does a student learn better with AI than without? In fact, I would venture to guess a student learns better with AI than if they had a personal tutor.
Rich people can afford personal tutors. They won’t do as well as people who have access to this AI. It’s a holistic kind of approach. I won’t go into much. It’s a pretty complex system. And I won’t go into the complexity of the system, but this is not just a chatbot. This is not just a key sort of use. AI simply. This has been built based on, and we’ve been working with Sarvam here in India to propose Diksha, which is a large collection of content in India, which is mostly unusable, to be honest, and build a 3 .0 version of Diksha, which is an AI -first experience. This is built on billions of student questions that have already been asked on the CK12 website.
Billions that is used to train the model to know how to teach a student. So it also has a teacher professional development curriculum, so teachers can keep up with it and keep up with most modern education. Again, compatible with the national education standards in India and the CBSE curriculum. Before. I go talk about AI doctors, I’m going to make a couple of comments. the AI tutors I’m talking about are far superior to human tutors. Here’s what they can do. They can quickly assess a student, where they are, in minutes, 10 minutes or 15 minutes, and then teach a tutor to the gaps in what the student doesn’t know through a complex process called knowledge tracing or tracing what they don’t know.
Moving on to AI doctors, which are also entirely possible today. In fact, these will make available 24 -7 to every Indian for almost trivial or no cost. Full primary care expertise, full disease management, chronic disease in India has been going up very, very dramatically. free mental health therapy, free physical therapy, and health and nutrition coaching. A level of comprehensiveness in AI health that isn’t available to the people who have the highest, most best -paid doctors in the world. None of this is available at this level, even in the U .S. or most Western countries. More than that is possible here for almost no cost. And of course, these AIs will be smart enough to know when to triage up to a human to do whatever functions only humans can do.
But let me not delude you. There’s very little a human doctor can do that this AI can’t do today. Other than the physical parts. If they have to feel your stomach, of course an AI can’t do that yet. I also fundamentally believe these services, AI -based doctors and AI -based personal tutors, should be part of the Aadhaar system. Aadhaar allowed us to offer UPI. There’s no reason we can’t offer on the same identity -based system where the hard work has already been done within a year or two to every Indian these services. So what I’m specifically proposing, to build a Section 8 nonprofit company to build, operate, and transfer into the Aadhaar ecosystem such systems. I think they’re relatively simple to do.
They need many cycles of iteration to adapt specifically to Indian conditions, all the Indic languages, all the differences in diseases. in each part of India. So I’m very, very excited about this. And what you can do for not only education, healthcare, the third element I want to talk about is agronomy. Having every farmer have a PhD level agronomist available locally 24 -7 alongside a UPI -like service as part of the Aadhaar system. A tutor that can engage students, it can find and teach to Gapson students and make the current Diksha system much more useful, much more friendly and leverage all the great content that is in the Diksha system in India today, but is not really usable because there’s no way to organize it and have an AI tell you what part of this vast system this vast library is relevant to you.
on the day you’re trying to do your homework or prepare for a test. You can multiply India’s doctors’ resources. So many years ago, I looked at the question of how you could scale the doctor -patient ratio in India to the same level that is in the West, like in the United States. And it wasn’t possible, even if you had a trillion dollars and decades to do this. That’s how far behind we are. But this will get us in India an opportunity to get well far ahead of the level of care available in a country like the United States, at least at the doctor level. There’s still surgeries. There’s still drugs. Those are separate matter, all areas in which AI can help.
But that’s sort of my hope. So the AI talks directly to patients. It diagnoses, prescribes tests, prescriptions. We are using technology from a company that’s been developed over the last five years and with Sarvam’s help and adaption to Indian languages using the Sarvam model. And you start with physician approval of the dialogue. So you oversee the AI with the doctor initially for the first couple of years. So think of AI as an intern, fresh graduate, an MBBS graduate who works for the doctor. They let them do a lot of things, but then they oversee them and watch that. And that’s the model I propose is possible in the next year or two. And within two or three years, I think that need for supervision will go away.
And of course, there’s emergencies. Sometimes you have to send somebody to the emergency room or the hospital, so AI can do that. The same is possible with agronomy, the third service. I won’t go through the details of this. I will try and finish up here. But I want to finish by saying the future is here today. Today, these massive impact services that couldn’t be done with hundreds of billions of dollars can be done very, very cheaply. Scale medicine, scale teaching, scale education, scale agronomy. And these services impact the bottom half of the population more, and they need it more than almost anybody else. That’s exciting. If we don’t do that, it is a massive opportunity loss for us.
Thank you.
“CK‑12 already provides free AI‑generated content to millions worldwide, with about four million Indian users and more than twelve million regular users.”
A source notes that roughly four to five million students in India have accessed CK-12 tutors, supporting the reported user numbers [S53].
“AI‑driven primary‑care doctors could handle diagnosis, test ordering, prescriptions, chronic‑disease management, mental‑health therapy, free physical therapy, and nutrition coaching.”
The discussion references the idea of AI primary-care and AI doctors, aligning with the claim that such a system is being envisioned [S53].
“Khosla’s view that AI‑driven impact must first reach the lower‑income half of India’s population to generate large‑scale change.”
Other sources emphasize the risk that public-sector adoption is needed for the poor to benefit from AI, providing broader context to this claim [S22].
“The AI‑tutor platform will be delivered as a “Diksha 3.0, AI‑first experience” and embedded in the Aadhaar ecosystem.”
While the knowledge base does not confirm these specifics, it discusses the importance of integrating AI tools with national digital infrastructure, offering relevant background [S8].
“Khosla described AI tutors as potentially outperforming human tutors and enabling better learning than private tutoring.”
Multiple sources highlight AI tutors as a way to democratize education and provide scalable tutoring, supporting the general premise though not the performance comparison [S14] and [S52].
The discussion shows a clear convergence on viewing AI as a powerful, positive catalyst for large‑scale social change. Both speakers adopt an optimistic tone—Speaker 1 about job automation, Khosla about AI‑enabled education, health and agriculture—suggesting a shared belief that AI’s challenges can be turned into opportunities. However, agreement is limited to this broad framing; detailed policy or implementation specifics are only advanced by Khosla, with no direct counter‑points from Speaker 1.
Moderate consensus on the overall optimistic narrative of AI’s impact, but low consensus on concrete strategies or sector‑specific proposals.
The brief exchange shows virtually no direct conflict. Speaker 1’s introductory comment about AI‑driven job displacement and Khosla’s detailed proposals for AI tutors, doctors and agronomists are complementary rather than contradictory. The only point of divergence is the focus of their optimism—employment versus inclusive service delivery—but both agree on AI’s large‑scale transformative potential.
Low. The lack of substantive disagreement suggests that the participants are aligned on the overarching goal of leveraging AI for development, which may facilitate consensus‑building on implementation pathways.
Vinod Khosla’s remarks repeatedly introduced fresh, high‑impact ideas—AI tutors, AI doctors, AI agronomists, and their integration with Aadhaar—that shifted the discussion from abstract optimism about AI to concrete, equity‑focused solutions for India’s largest unmet needs. Each pivotal comment opened a new thematic strand, deepened the analysis by confronting feasibility and safety, and reframed the narrative toward inclusive, large‑scale transformation. Collectively, these insights steered the conversation toward actionable policy and implementation pathways, turning a generic keynote into a roadmap for leveraging AI to serve the bottom half of the Indian population.
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
