From being a buzzword to becoming a key technology, we see AI moving forward in India. But, one question remains - are there enough funds to benefit the long-term value of AI?
First there was human effort. Then came the machines. And then computers.
Over the last few decades, from being used for small tasks like calculations and data storage, computers have accelerated economic and social growth. But going forward, the next big thing will most certainly be artificial intelligence (AI).
“Algorithms that can increase Return on Investment (RoI) by understanding consumer behaviour are the future of the industry. Banks, retailers, and consumer goods companies are all looking to understand their audience affinity to drive up sales,” says Aditya Bhamadipaty, CEO and co-founder of First Hive, selected as one of YourStory’s Tech30 companies for 2018.
He says there is no stopping AI-based algorithms as they can crunch large volumes of data without human intervention. The writing is on the wall - the era of AI is here, and here to stay.
Increasing investments, but what about long-term plans?
And what better measure of this than investors’ confidence in the sector. According to YS Research, when compared with 2017, when only $8.8 million was invested across in 17 deals in AI-based companies in India, as of October 2018, a total of $297 million had already been invested in the sector across 25 deals this year. One may wonder why this has happened.
“We see a lot of noise in this space, but there are Indian entrepreneurs who are solving some key issues in several industry verticals,” says Manu Rekhi, Partner at Inventus Capital.
Being the second most populated country in the world - and the fastest growing economy - it's not only India that needs AI; AI also needs India to grow. But the bulk of investments in this space took place elsewhere. Among the largest funding rounds in the AI industry this year were Dataminr ($392 million in a Series E round) and CrowdStrike ($200 million, also in a Series E round). Both companies are based in the US, which continues to dominate AI funding with a total of $21.3 billion raised in the first six months of 2018 with major investments in security and threat responses.
But, as data shows, investments in AI are increasing in India. The next question is - are there enough funds to benefit the long-term value of AI? Among the large investments this year, Bengaluru-, now Palo Alto-, based ThoughtSpot (AI for business intelligence) raised $145 million while Hyderabad-based ThinCL (AI based processors) raised $65 million.
“If one really has to ensure that AI has to take off in India, there has to be a strong math-based learning system for teachers and students across the country, not just in top schools. AI should be taught as a subject in colleges and students should be encouraged to build things in the open source community,” says Manan Khurma, founder of CuMath.
Decoding AI, sector by sector
As of now, AI as a field and its applications have not yet been entirely understood by all, and interest in the sector will only come about from understanding how it can impact lives. Take for example Pune-based Fluid Robotics. The startup makes robot crawlers to clean up clogged drains, making it unnecessary for a human being to enter drains. With Fluid Robotics, the job of the drain worker is that of operating a robot that cleans drains and collects data on clogged pipes for efficient city management.
Another example could be radiology. An average radiologist scans over a thousand images a day, making human error a possibility. Bengaluru-based Actify Data Labs is building a platform that can help radiologists use AI to scan over 5,000 images or more a day. Here, AI brings down the cost of healthcare, and can nearly eliminate any human error. “Only AI can crunch such volumes of text and images. With so much digital data, there is no way a human can comprehend even a percentage of the digital explosion,” says Hindol Basu, founder of Actify Data Labs.
Then there is Jubi.AI, which is changing the way financial products are sold to customers. Jubi’s conversational bots help call centre workers meet screened applicants instead of cold calling applicants. “Our bots have been trained to have real conversations with people looking for advice on investments in financial products. It’s better curated than talking to human beings,” says Subhadeep Bhattacharya, Co-founder and CEO of Jubi.AI.
AI disrupting traditional sectors
Even when it comes to older industries like construction, AI is showing it has a role to play. Builders like Kolte Patil Developers and Purvankara have taken to working with startups like Racetrack.AI to increase the effectiveness of their marketing teams. Racetrack uses machine learning to reduce customer waiting time by half, and helped the builders increase conversion from two percent to seven percent.
Subrat Parida, Founder and CEO Racetrack.ai says, “This is an overwhelming and positive moment for us. We are now looking forward to and are excited to announce our next overseas operation base in the US, Mauritius, and Singapore.”
There are misconceptions, too. Most people would say any software that makes recommendations based on buying patterns is AI, or any software that has the ability to learn and suggest intuitively is AI. That is not true.
AI learns based on several patterns. Using complex applications like mapping genome data and the daily routine of a person AI can predict likely diseases and suggest changes for a healthier life. The key to this is the robustness of the data and the algorithms that make up the AI workflow.
Government thinktank Niti Aayog this year announced it would form an AI policy but it is yet to influence a sea change in AI adoption. A talent scout told YourStory that there aren’t more than 100 qualified AI scientists in India. China and the US claim to have at least half million such experts each. A Tencent Research Institute report says there are only three lakh AI engineers worldwide, and there is a growing trend of students taking AI and machine learning courses. “Many students are taking AI courses online, outside of school curriculum, and we can help Indian engineers come to speed with AI,” says Ishan Gupta, Managing Director, Udacity India.
“Most Indian startup businesses will ultimately fly to the US because the market is in the US,” says Inventus’ Manu Rekhi. Startups should learn to scale up algorithms, and not just scale with the moniker ‘AI’ attached to their name.
Eventually, how we handle artificial intelligence to combine the work of an engineer and a sociologist will impact the learning ability of algorithms and make sense for the business value they generate.