Vir Das’ recent post highlights the bittersweet reality of an IIT PhD turned chauffeur in the US, shedding light on how India’s brightest minds struggle abroad—and fail to find support to thrive as startup founders at home.
Meanwhile, the researchers in the West have successfully built billion-dollar AI startups, thanks to its strong ecosystems, abundant capital, and supportive culture.
Vishnu Vardhan, the founder of Vizzhy and SML, aptly described the sad state of Indian AI startups in an exclusive interview with AIM, stating, “Indian startups often focus on business applications instead of foundational innovation, with investors prioritising quick returns over long-term deep-tech investments.”
He said that the so-called deeptech investors clearly have no theses whatsoever. “I met a few VCs who said they were deeptech investors. I asked them their ticket size; they don’t even understand the scale of investment required for true deeptech.”
Vardhan said that India has no foundational LLMs despite having some of the best engineers and researchers because both industry and government fail to invest adequately in research— “$10–20 million budgets pale compared to MIT’s $6 billion annual funding.”
He added that the IT industry has thrived on a service model, but building foundational AI requires factories, not just manpower, and the mindset hasn’t shifted.
There is still hope: Ganesh Ramakrishnan, professor at IIT Bombay and co-founder of bbsAI, is addressing the sad state of Indian AI research—often hindered by limited data, lack of investment, and reliance on foreign models—by developing deterministic, small language models and pioneering native language technology for inclusivity and innovation.
“In India, only those who know English have privileged access to technology. Most developed nations have access to technology in their native languages,” shared Ramakrishnan.
“In R&D, my focus is on innovation, but I would prefer to invest more in the ‘D’ (development) than the ‘R’ (research),” said Arjun Rao, General Partner, Speciale Invest on how he thinks basic research can be funded by academia, government grants, philanthropy, or corporate social responsibility initiatives. This is reflective of a broader VC sentiment in India with respect to funding researchers in the country.
That explains why India still doesn’t have something like Mistral AI yet. Although India’s incubator culture has made significant strides, more efforts are needed to support traditional researchers in transitioning into founders.
“I don’t think there is any point in replicating what’s out there (like Mistral). And there’s no point in training yet another large language model,” said Paras Chopra, founder at Turing Dream, an AI residency for developers and researchers in Bengaluru.
He further highlighted how future breakthroughs in AI would come from going beyond replication – exploring new concepts and possibilities. His projects named Blueberry and Tofu are focused on exploring new dimensions of AI, beyond just product development.
Conversely, regions like Silicon Valley offer networks of investors and incubators. While the acceptance of failure encourages risk-taking, incubators like Y Combinator solidify the West’s prowess in AI innovation.
Interestingly, OpenAI was originally created at YC Research, before Elon Musk and Sam Altman chose a different path. As of early 2024, YC has backed over 4,500 startups worth $600 billion, solidifying its role as a key resource for new founders, including those with strong academic backgrounds.
However, there are cases where researchers have not made the cut either. In some cases, founders have also returned back to big tech after launching their own startups.
Big Tech Defectors
François Chollet, the creator of Keras, recently announced his departure from Google. Not long after, Toby Shevlane, a scientist at Google DeepMind, also revealed that he was leaving the company to pursue his own venture. This trend is not new as there is a shift towards researchers moving from academia and research labs into entrepreneurship as startup founders.
These departures in the AI industry often involve researchers leaving big companies to start their own ventures for more freedom, focusing on ethical and societal challenges in AGI development, or launching startups centred on open-source AI.
The concept of ‘innovator’s dilemma’, introduced by Harvard professor Clayton Christensen in his book, highlights why successful companies struggle to adapt to disruptive innovations despite their capabilities and cause people to leave.
Nearly all researchers who co-authored Google’s transformers paper ‘Attention Is All You Need’, which shaped the foundation of modern AI, later left the organisation to start their own companies to tackle specialised niches within AI.
“…At that time GPT-2 had just come out and the trajectory of the technology was pretty clear…So I called up my co-founders and I said ‘Maybe we should figure out how to build these things’,” said Cohere CEO Aidan Gomez in a recent podcast, elaborating on the need to capitalise on the wave future of internet models.
Even after ChatGPT’s launch, several OpenAI researchers left the company to start their own ventures. These usually involved either building proprietary software or applications on the current software. Close to 75 employees have left OpenAI and founded around 30 AI startups, which have been growing since.
Research Roots
The rising demand for AI solutions presents significant opportunities for researchers and early investment prospects for VCs, comparable to the early stages of Amazon, Google, and Facebook.
Tech giants like Google, Microsoft, Amazon, and Facebook all began as research ideas, ranking websites, programming tools, algorithms, and network theories, thereby shaping innovations used by millions today. Leading tech companies either acquire or heavily fund research firms to close the gap in their AI advancements.
We see a similar pattern with AI startups, especially those that have originated in research labs or were founded by former researchers.
For instance, DeepMind was founded in 2010 by Demis Hassabis, Shane Legg, and Mustafa Suleyman. Even before Google reportedly acquired it for $400 million, DeepMind raised significant funding and introduced technologies like deep learning and reinforcement learning.
Anthropic, founded by former OpenAI researchers, focuses on advancing AI safety and developing large-scale language models. Amazon’s $4 billion funding of Anthropic reflects its interest in betting big on the AI wave, despite not having an initial lead in this field.
Startups like Cohere, Hugging Face, Stability AI and Mistral AI have similar backgrounds.
Billion Dollar Club
The market opportunity for AI is already huge and will only increase. As per reports, the AI market is projected to reach $780–$990 billion by 2027, growing at an annual rate of 40-55%.
During an interview, PayPal co-founder Peter Thiel referred to investments in AI as still fraught with danger, comparing the state of AI in 2024 to the internet in 1999, highlighting the coexistence of immense potential and substantial risks – especially in the long term.
Moreover, investors are attracted to founders due to their strong technical footing, making it easier to secure venture capital. In this respect, AI researchers are uniquely positioned to build million, or even billion-dollar startups.
Fei-Fei Li, co-director of Stanford’s Human-Centered AI Institute, co-founded World Labs in 2024. The institute, valued at over $1 billion, aims to develop AI systems with advanced spatial intelligence for 3D interaction.
Similarly, Ilya Sutskever, co-creator of ChatGPT, shifted his focus to AI safety by founding Safe Superintelligence (SSI), reportedly valued at $5 billion. Meanwhile, computer scientist Kai-Fu Lee in March founded 01.AI, valued at $1 billion, to develop open-source LLMs specific to China.
In the end, being a founder, traditional or otherwise, requires the right skills and determination.
Reflecting on her own educational journey at AMD’s event in Bengaluru last week, AMD CEO Lisa Su said, “This is a once-in-a-lifetime thing, the opportunity to get a PhD. But education is definitely not doctrine. The most important thing is the process which we have gone through and learned from.”
The post Indian AI Researchers Can Only Dream of Building Billion-Dollar Startups appeared first on Analytics India Magazine.
