Frontier-tech: Are we there yet?

by RohitAug 10, 2017 | insert_comment

During a recent meeting with a prospective LP, I was asked about whether India is capable of producing breakthrough technology companies. The point being made was that over the years Indian technical education system has dumbed down to mass-produce recruits for IT services companies, and is now lagging far behind countries like China, US, and Israel in cutting edge research.

I offered evidence of some fine startups that were building good technologies, but I think the larger point really is indisputable — by any yardstick, be it % of GDP spend on research (India spends .8% compared to 2+% for all major economies) , global ranking of Indian universities (none in top 50 and only 3–4 in top 200 compared to China with 4 in top 50 and several more in top 200), papers published in top journals and conferences, India lags far behind most developed countries, and emerging powerhouse China. Startups rarely have the financial muscle for ground-breaking research. Rather, they are created by scientists and engineers that are product of long years of investments in universities and govt funded or corporate owned research labs. In the US, it’s the GAFA — Google, Apple, Facebook, Amazon — that have used their ginormous cashflows for industry disruptive initiatives in areas like AI, driverless cars, robotics, and life sciences. The steady movement of people from these companies to new startups, and furious pace of acquisitions and investments make it an environment that nurtures and rewards talent for pursuing deep technologies. Similarly, in China, the BAT trinity (Baidu, Alibaba, Tencent) have equally ambitious efforts of their own in the field of AI.

In India, both the academia and the corporations are clearly lagging behind. It will take a sustained push on institutional capacity building on multiple fronts to bridge the gap and it will not happen overnight. China recently announced a moonshot mission to be the leader in the AI in the world by 2030. Perhaps India needs its own moonshots to drive this.

Where do we go from here?

It is clearly premature to jump on the ‘frontier-tech/deep-tech’ cheerleading bandwagon yet. But while being a realist about the current situation, I’d argue that the environment today is fundamentally more supportive for Indian startups to participate in the deep technology revolution because of a number of factors.

  1. Open Source: Unlike space and defence era research that was closely guarded by the governments, important advances today in fields like Deep Learning are being made first for consumer products by companies like Google and Facebook. The research is openly available and so are the frameworks (TensorFlow, Torch) that make it easy for new developers to build on.
  2. Cloud: Cloud computing has now become so ubiquitous that we take it for granted. But it was not so even a few years ago. I remember the time when I was an undergrad student at IIT -D doing a project on HPC. I was lucky to have access to a Param class supercomputer in the lab, but once it developed a hardware snag mid-way though the project, there was no alternative. Now anyone can spin up a powerful supercomputer on-demand on platforms like AWS and Google Cloud. It’s easy to overlook how big a leveller cloud has been. 
  3. Hardware supply chain: The scale of mobile phones revolutionised the hardware supply chain like never before. The scale made it cheap enough to prototype and the small form factor and variety of sensors makes it possible to measure any phenomena. This is what has really brought the tech into daily life creating all kinds of possibilities from drones to health bands. 
  4. Open Learning: Another exciting transformation that has happened is that education is no longer confined to classrooms. Platforms like Coursera and NPTEL (offered by IITs) have made quality material accessible to everyone. Forums like Kaggle provide a practice ground for solving real-world problems. There is still no replacement for the learning that happens when you engage with a bunch of smart people in a room, but for the motivated learner, these platforms provide a great start.
  5. Ecosystem: In the last decade, a steady stream of Indians with advanced degrees and experience in tech companies in the US has been coming back to India. This is a rich talent pool that acts as a bridge for transfer of knowledge and best practices from top tech companies to local startups. Almost all large global companies today have an R&D centre in India, and in last couple years, they have started engaging with the local startup ecosystem. Companies as diverse as Airbus, SAP, Target, and SwissRe have opened accelerators and some have started investing through their venture arms. It is nowhere close to the action in the Valley but it’s a promising start.

So even though the Indian ecosystem is behind when it comes to laying the foundation of deep technology, the opportunities to build on top of that stack and create global tech companies are present and clear. And that’s what I am excited about. In my next post, I’ll cover how companies need to think beyond just the tech, and focus on products and solutions needed to create value for the customers