AI dilemma: To be or not to be

In May, SenseTime Group, a Chinese technology company specialising in artificial intelligence (AI) and facial recognition, raised $620 million in a funding round led by Fidelity International, Hopu Capital, Tiger Global Management and Silver Lake Partners. The three-year-old startup, arguably the world’s most valuable startup in AI, was valued at $4.5 billion. The funding came at a time when the Chinese government is pulling all the stops to make the country a global leader in AI over the next seven years. By 2025, the Chinese government intends to grow AI industry’s value over $60 billion.

In India, the picture is different. Investors in India still seem wary of writing big cheques for home-grown AI firms. A large percentage of cheques are less than $5-$7 million. Earlier this month, Racetrack.ai, which provides AI for sales support and customer issues, raised $5 million in its pre-series A funding from investors, including Biocon founding member Murali Krishna, and Dov Kagan, a U.S.-based investor. A few days ago, Bengaluru and San Francisco-based AI startup Actionable Science raised $1 million in a seed-funding round from Globevestor and Tri-Valley Ventures. Last month, Bengaluru-based artificial intelligence startup Dishq, which leverages food science and machine learning to predict people’s taste, raised a pre-seed funding of $400,000. In the same month, emotion recognition startup Entropik Tech raised $1.1 million led by BIF (Bharat Innovation Fund) along with IDFC-Parampara Capital, Arthavida Ventures and Jitendra Gupta (managing director, PayU), among others.

According to Tracxn data, between 2014 and 2017, AI startups in the country raised less than $100 million from investors. Why?

There is no lack of unique, innovative models, says Debjani Ghosh, president of IT industry body Nasscom, adding that investors in India are not ready to take big risks. “The power of investing is when you back an idea that is crazy… it’s ahead of its time. If we (investors) continue to invest in things that are proven, how will we innovate or drive scale,” she says.

AI is an essential part of every tech firm’s strategy, so much so that billions of dollars are being invested in these startups world over (with the U.S. and China leading the charts) and global biggies like Facebook, Amazon and Google have been on an acquisition spree. Interestingly, some of the merger and acquisition targets have been in our backyard. In 2017, Google acquired four-month-old Bengaluru-based AI startup Halli Labs for an undisclosed amount. Chinese mobile Internet company APUS Group acquired SIFTR Labs, an AI-driven photography curation platform.

However, a significant proportion of venture capital (VC) investors in India still prefer to back models that have a proven track record of success in other parts of the world, such as those in online retail, education or healthcare. Also, they want to back a model where they can see a billion-dollar opportunity of growth and scale. AI startups in India are not showing that opportunity just yet, says Anil Joshi, managing partner at Unicorn India Ventures, a VC fund. “It (AI) can never be a $1 billion company; it can help companies generate billions of dollars. AI will always work in the backend, it’s an enabler. It’s like this—you know a problem and you add a layer of artificial intelligence to solve it. Hence AI (startup) cannot be an Ola or Uber, it’s an enabler at the backend,” he says.

Also, there are other constraints. Finding the right professionals with the best acumen for upcoming fields such as AI and analytics is not easy. Also, they cost a lot more. “Unless one is a subject matter expert, building AI solutions is not easy,” says Joshi, adding that currently ecommerce and education are two sectors that are leveraging AI in a big way in India.

Also, unlike companies in the U.S., a majority of people in India are yet to understand the real value of data and data analysis. Most big Indian companies are not comfortable sharing their backend operations with others, says Joshi.

Joshi is not alone in pointing out these constraints. While the government think tank NITI Aayog identified healthcare, agriculture, education, smart cities and infrastructure, and smart mobility and transport to highlight the potential of AI tools and technologies in transforming these sectors, it also outlined a few challenges. According to a discussion paper released in June, these hurdles include: lack of enabling data ecosystems; low intensity of AI research; inadequate availability of AI expertise, manpower and skilling opportunities; high resource cost and low awareness for adopting AI in business processes, and unclear privacy, security and ethical regulations.

It may not be easy but AI startups are today crucial for companies. Take the example of e-commerce firms, probably the biggest users of AI. If one online retailer uses AI better than its competitor, it can almost ensure that its customer does not go to the competitor as it will keep throwing up options for the user that are to his/her liking. Joshi says AI can boost the return on investment (ROI) of digital marketing by as much as 40%.