Comment by Riaz Haq yesterday

Sandeep Manudhane
@sandeep_PT
Bro, we all support you, and speak highly of you and Sarvam now, but refrain from making claims that don't stand technical scrutiny. Sarvam is no DeepSeek. There has been no new foundational algorithmic breakthrough, no training on fully indigenous chips, and no step-change in the economics of frontier-scale training. DeepSeek’s low-cost result came from concrete (new) technical advances, novel model architecture choices, deep bare-metal optimization of their hardware stack, and access to massive, diverse global data at scale. Let's not confuse running a smaller regional cluster with executing a global paradigm shift.
We wish you the best.
#Sarvam #DeepSeek #AI

https://x.com/sandeep_PT/status/2025097332208091309?s=20

----------


Post

See new posts
Conversation
Sabeer Bhatia
@sabeer
The Galgotia fiasco is the inevitable end result of what I said about education a while back. When institutions prioritize optics, money, and rote memorization over real learning and critical thinking, failure isn’t an accident - it’s the outcome. https://economictimes.indiatimes.com/magazines/panache/hotmail-cofo...


https://x.com/sabeer/status/2024879694701797589?s=20

-------------

Dr Nimo Yadav 2.0
@DrNimoYadav
Galgotias University is the only university in the world that published a research paper claiming that coronavirus gets killed by banging thalis and bells.

And Modi awarded this as Top Private university of India.

God save this country 🙏🏾

https://x.com/DrNimoYadav/status/2024696987820708102?s=20

--------------


Prateek Goyal
@tweets_prateekg
Fun facts:
In just 7 yrs,#Galgotias Universty says it has filed 2,430 patents.

Meanwhile,IIT-Madras,one of India’s most respected public research institutions has filed about 2,550 patents in the 50 yrs since its inception in 1975.
someone needs to patent the word "patent". 1/n

https://x.com/tweets_prateekg/status/2024755147126255854?s=20

-----------


Mint
@livemint
Galgotias University was allotted more space than four IITs' total during the AI Impact Summit 2026. This sparked controversy over the space allocation process, questioning the criteria used to determine exhibitor space at the event aimed at showcasing India's tech advancements.

https://livemint.com/news/india/ai-summit-galgotias-universitys-boo...

https://x.com/livemint/status/2025053947745153137?s=20

Comment by Riaz Haq yesterday

India Can’t Spectacle Its Way to AI Power

https://www.bloomberg.com/opinion/articles/2026-02-19/india-can-t-s...


Giant posters of Prime Minister Narendra Modi and feel-good slogans about artificial intelligence lined New Delhi’s roundabouts for the AI Impact Summit.
The summit was marked by dysfunction, including road closures, long lines, and unexpected venue closures, but also revealed India's appetite for AI and its potential as a market.
India's AI adoption figures are high, but the country faces challenges in moving from being a consumer to a producer of AI, including finding land, water, and electricity for data centers and addressing environmental concerns.


The spectacle of this week also obscures a more immediate, unique risk. Throughout the developed world, policymakers sell AI as a solution to aging populations and labor shortages. India has the opposite problem: a huge, young, increasingly educated workforce that needs jobs. The recent “AI scare” selloff has hit Indian IT especially hard, a reminder of how exposed its software sector is. If AI becomes a substitute for entry-level work before India can generate new opportunities, the social impact could be sharper than in the countries exporting the technology. The challenge of translating the AI wave into livelihoods is a bigger governing test for Modi than collecting selfies with Silicon Valley’s elite.

Still, the enthusiasm at the summit wasn’t manufactured. I spoke to a couple of college students who came to check it out on Monday. They didn’t mind the crowds or disruptions, saying the chaos simply proved how much people cared. They insisted that Indians aren’t merely using AI, they’re experimenting and building. The optimism was contagious, and it hinted at the nation’s genuine advantage: a massive, ambitious, mobile-first talent pool willing to try new tools fast.

But optimism doesn’t substitute for an ecosystem. A harder question hanging over this gathering is why India, with undeniably deep tech talent, has never had a “DeepSeek moment,” and still lacks a defining foundational research breakthrough. Adoption can scale quickly, but it’s much more difficult to build frontier capability without sustained investment in research, access to compute, and the kind of capital that lets entrepreneurs take bold bets. If the summit was meant to showcase India as an AI builder, the disarray also exposed why so many of its best and brightest are seeking opportunities elsewhere.

AI’s promises and hypocrisies were on open display in Delhi. Under a banner of “democratizing AI,” hotel rooms went for as much as $33,000 a night while homeless people were forcibly moved along the road to the venue. India is a test case for whether AI diffusion empowers everyday people or widens inequality. The rest of the world will be watching closely.

Walking between meetings in downtown New Delhi, I stopped counting the number of Modi posters after I hit 20. India can host the world. It can sell its vision of the future. But it can’t spectacle its way into AI power. That takes the unglamorous work of dedicated research funding, trustworthy institutions, reliable infrastructure — and a plan for the people expected to live with the consequences of this tech revolution.

Comment by Riaz Haq yesterday

Bloomberg Opinion

@opinion
Nowhere were the promises and hypocrisies of AI clearer than in India this week.

@cathythorbecke
explains from New Delhi 🎥

https://x.com/opinion/status/2024918460430455010?s=20

Comment by Riaz Haq yesterday

Hotmail cofounder Sabeer Bhatia blasts Indian education system: 'We are producing an army of useless kids'

Read more at:
https://economictimes.indiatimes.com/magazines/panache/hotmail-cofo...

Synopsis
Hotmail co-founder Sabeer Bhatia criticizes India's education system for prioritizing obedience over critical thinking, leading to a workforce of unoriginal individuals. He contrasts this with Western education, which encourages independent thought. Bhatia argues that the Indian system stifles creativity by emphasizing rote memorization and discouraging failure, hindering innovation and producing compliant workers instead of visionary creators.




In a nation obsessed with ranks, marks, and job titles, we’ve forgotten to ask the most basic question: Are our kids actually learning? For Sabeer Bhatia, the man who co-founded Hotmail, the answer is a resounding no. In a hard-hitting interview on the NNP podcast, Bhatia doesn’t mince words, calling out India’s education and work culture for building an “army of useless kids” instead of original thinkers.

At the heart of the problem is a deeply flawed system that rewards obedience over curiosity. “We live in a conformist society—people are often told, ‘Listen to others, do what they say.’ But why follow a path that’s already been walked?” Bhatia questions. Whether it's memorising textbooks or chasing the same startup ideas in saturated sectors, he believes the country is wired to produce workers who take orders, not visionaries who disrupt systems.

Bhatia draws a stark contrast between Indian and Western education. His young children in the US write their own stories and ideas, even if full of spelling mistakes. “Teachers don’t correct those because spelling is irrelevant. What matters is the thought.” In India, however, children are punished for errors instead of being encouraged to think independently. They’re taught not to learn, but to score.

It’s a mindset that starts early and ends up defining careers. Many bright students become engineers or doctors, not because of passion, but because society deemed it safe or respectable. “You can’t suppress the arts, sports, and culture and expect to build a balanced society,” he adds.

And even when Indian youth show entrepreneurial ambition, Bhatia says they’re crippled by the system itself: “You’re never asked to write a paper. You’re asked to memorise 13 chapters and regurgitate them. That is not education.

For innovation to thrive, he insists we need critical thinkers—people who do things, build things, experiment and fail. But the stigma around failure in India is so strong, even someone like Bhatia has been asked, “What have you done since Hotmail?” As if one stumble erases all worth. He believes that until India stops confusing compliance with intelligence, the country will keep losing potential to a system designed for factory workers, not creators.

Comment by Riaz Haq 21 hours ago

India Has Its Own Unique Recipe for ‘AI Sovereignty’

While no country can fully decouple from US or Chinese models anytime soon, alternatives are emerging.

https://www.bloomberg.com/news/newsletters/2026-02-21/india-has-a-u...

Everyone at the AI summit, whether from India, elsewhere in Asia, Europe, Africa or Latin America, likely understood one inescapable fact: They’re all at the mercy of two countries — the US and China — and a small handful of companies that control AI technology.

This, fundamentally, is a matter of sovereignty: Whether a nation’s AI systems can be independent of foreign authority. That danger was showcased in 2024, when members of Australia’s UniSuper pension fund had access to their accounts cut off due to a Google cloud misconfiguration. In October, Amazon.com Inc.’s cloud services — the world’s largest — also suffered a major shutdown, damaging its reputation.

While such incidents were declared accidents, they opened eyes to the possibility the next one might not be.

“We know ‘open’ can become ‘closed’ at any point in time — the moment a CEO decides,” Ashwini Vaishnaw, India’s minister for electronics and information technology, said at a Bloomberg New Economy breakfast on the summit’s sidelines.

That could happen for any number of reasons — a change in corporate strategy, a commercial dispute or a government order.

Even Mukesh Ambani, the billionaire Indian industrialist who ranks as Asia’s richest man and is plowing money into AI, framed the implications of this new dynamic in binary terms.

“Will AI concentrate power in the hands of a few, or will it democratize opportunity for all?” Ambani asked in his Feb. 19 plenary address at the summit.

While no country can fully decouple from US or Chinese AI models anytime soon, alternatives to total dependency are emerging. One recipe came from Rishi Sunak, who as UK prime minister hosted the first AI summit in 2023. He’s now a senior adviser to Goldman Sachs Group Inc.

First, a nation could create its own leverage, like the Netherlands with ASML Holding NV, whose equipment no advanced chipmaking operation can do without. As Sunak put it, “Where can we occupy a very strong, ideally monopolistic position in that supply chain?”

Next, diversification. “Avoid vendor lock-in, have a mix of providers, make sure you’re building capabilities that will allow you to switch things in and out,” Sunak said. Add in partnerships — even in a world of geopolitical volatility, there’s room for allies to share knowledge and intelligence that enhance mutual AI sovereignty.

“If you do all three of those things, you can build a strategy that gives your country resilience but also the opportunity to take advantage of this incredible thing that’s in front of us,” Sunak said.

India, meantime, has a vision for AI that’s distinctly Global South. “Welfare for All, Happiness of All,” Prime Minister Narendra Modi proclaims on one of the many billboards in New Delhi commemorating the summit. “Showcasing the Power of AI for Global Good,” reads another.

The country, Vaishnaw explained, has its own multipronged plan to achieve AI sovereignty:

Support the local development of so-called small language models based on parameters critical to the Indian market.
Build a “common compute facility” so students, entrepreneurs and startups have affordable access to AI.
Design and ultimately manufacture its own semiconductors.
Use the AI summit to attract venture capital for Indian startups as well as funding for computing infrastructure such as data centers.
Before the summit began, Vaishnaw said India had obtained commitments for about $5 billion of venture capital investments and $140 billion for AI infrastructure. Days later, that had increased to at least $17 billion and $270 billion, respectively. “Not bad,” he mused.

And enough to bring India one step closer to realizing its grand ambitions.

Comment by Riaz Haq 3 hours ago

Sanjay Jha
@JhaSanjay
HOW WE MAKE A FOOL OF OURSELVES WITH PR : #AISummit

America is USD 30 trillion economy. China is at $20 trillion. We are officially not even $ 4 trillion yet. The gap is likely to widen even further.

America and China have social security architecture and higher standard of living for the average citizen .

Under Modi, our income inequality has worsened. 800 million live off free food grains. We are the lowest among G-20 countries in per capita income. On education, health, life expectancy, corruption, deaths on the road, environment, every conceivable parameter, we are awful. Our best are leaving the country, and many who are here, are suffering and jobless.

AQI related 2 million deaths happen every year. Zero action.

This sudden AI hype makes little sense. Sure, do what you should, but take a pause: Why is India becoming a public relations platform for the most failed leader in our political history???

https://x.com/JhaSanjay/status/2025428142920507532?s=20

---------------


vir sanghvi
@virsanghvi
After the AI Summit when citizens were treated like dirt because of the VIP culture this is The Economist’s take. It’s the right point: ordinary people should not be victims of VIP worship

India’s VIP culture is out of control (from
@TheEconomist
)

https://x.com/virsanghvi/status/2025549576225300694?s=20

----------

The Economist
@TheEconomist
The country’s leaders claim to be servants but act like masters

https://x.com/TheEconomist/status/2025522218290049399?s=20

-------------


Sagarika Ghose
@sagarikaghose
Indian students lack critical thinking because of the pressure of rote learning. ‘Docile engineers’ may find it difficult to think critically or hold power to account: says the profound Martha Nussbaum

https://x.com/sagarikaghose/status/2025441216603848865?s=20

Comment by Riaz Haq 3 hours ago

India’s AI ambitions hit limits at global summit

By Krishan Kaushik

https://www.ft.com/content/5c26f2f6-c857-407c-93fe-7f59aa88c8f4

India’s ambition to widen access to AI and play a crucial role in the fast-developing technology fell short this week, as the country continues to struggle to find its place in an industry dominated by the US and China. Prime Minister Narendra Modi’s government used the country’s hosting of this year’s Global AI Summit to push top AI groups, including OpenAI and Google, to open source models for specific social ends, such as healthcare, education and agriculture.

“Some countries and companies believe that AI is a strategic asset and should therefore be developed confidentially,” Modi said in his address. He added that the technology would “only benefit the world when it is shared.” The call comes as India has struggled to become a significant player in the AI arms race. Despite its huge tech talent pool and being home to global IT groups such as Infosys and Tata Consultancy Services, it is not a leader in developing large language models or creating products from the technology’s rollout.

The world’s most populous country wants to make sure that AI decision-making is not restricted to the US and China, as countries of the global south recognise the technology’s transformative power and seek faster adoption. But India’s push to widen access and introduce a framework for global AI governance was largely dismissed by Washington and the country’s leading tech companies.

Michael Kratsios, White House chief of science and technology policy, on Friday told attendees that the US government “totally” rejected global governance of AI. “We believe AI adoption cannot lead to a brighter future if it is subject to bureaucracies and centralised control,” he added. Instead India managed to secure a voluntary commitment for AI companies to share their data on how the technology is being used and the effectiveness of multilingual models. The Global AI summit was also a roadshow for India, as it tries to minimise AI’s threat for its largest white-collar employer, the $300bn IT services sector. The event saw it secure investment pledges worth $227bn, most of it involving the build-out of data centres.

But the summit was marred by dysfunction, with gridlocked streets and attendees waiting in long queues to enter the venue. It was also hit by several high-profile speakers pulling out, including Nvidia chief Jensen Huang and Microsoft co-founder Bill Gates.

Despite big data centre projects being announced, experts warned that India did not yet have the large-scale computing infrastructure needed to become a big player in AI. Alphabet’s senior vice-president James Manyika told the FT that the wider access India wants was limited because of the lack of infrastructure, especially in the global south. “The world has not got enough capacity . . . I think the scale of the capacity of the investments, I think, is something that is going to have to happen everywhere. It’s particularly acute in the global south, so I think there’s work . . . to do.” The fractured geopolitical environment also worked against India’s push for a strong regulatory framework, according to AI experts.

J Trevor Hughes, chief executive of Boston-based not-for-profit IAPP, said the summit was conducted at an “odd moment” because “geopolitics is changing around the world”. “There is a broad deregulatory mood in the air. So the idea of imposing AI regulation creates an allergic reaction right now in many, and yet, risk management in AI is still a critical thing,” he added. A person who was part of the discussions between India and some of the US tech giants noted that as there was no pressure from the US government for global AI regulation, saying American companies are “feeling absolutely no need to even agree to a baseline”.

Comment by Riaz Haq 3 hours ago

India’s AI ambitions hit limits at global summit

By Krishan Kaushik

https://www.ft.com/content/5c26f2f6-c857-407c-93fe-7f59aa88c8f4

The summit showed there were signs that India was making some headway in AI. Sarvam AI, one of India’s leading AI start-ups, used the summit to launch its new LLM. The Bengaluru-based group’s model is focused on solving day-to-day concerns, rather than harder problems such as advanced math and complex philosophical questions.

Meanwhile, India’s IT services sector has belatedly decided to embrace the technology. TCS and Infosys both announced partnerships with OpenAI and Anthropic, respectively, to help their clients adopt and integrate the technology. The importance of the Indian market to the leading AI groups was also made clear. OpenAI’s Sam Altman said the country was the company’s fastest-growing market for Codex, its coding agent that works alongside developers to build software. But one of the most talked-about moments at the summit came on Thursday, when a group photo of industry and government leaders neatly showed the friction at the heart of the AI industry. While everyone in the line followed Modi’s encouragement to clasp their neighbours’ hand in celebration above their heads, two of the industry’s fiercest rivals, Altman and Anthropic’s Dario Amodei, who were standing next to each other, could not bring themselves to do so.

Comment