For our next interaction on this month’s theme “How will AI affect humans’, we caught up with Harshal Dhir, CEO and co-founder of Machaao.Inc. He has built Cricket.AI that offers near real-time Cricket score at personalised intervals. A technical leader with a great product sense, Dhir has over 18 years of experience in building consumer and enterprise systems for Fortune 500 clients. With a strong inclination towards building game-changing products that impact human lives in a positive way, Dhir along with his team has built a conversational prediction platform where sports fans can compete to win daily cash, brand rewards.
Analytics India Magazine: What are some of the practical implementations of AI that have revolutionised the way humans are functioning?
Harshal Dhir: We are currently living in a world surrounded with “first generation AI agents” built by companies such as Facebook, Apple, Google, Amazon, Uber that constantly collect and analyse data. These are being currently used for a variety of tasks such as providing contextual options in the real-time to being extensively used in the financial markets to predict market trends, along with various other metric analytics. I believe that the second generation of these AI agents, which as currently under development, would act on the data that is being collected and become more personal, autonomous and reactive in operation. I personally think we should be able to start witnessing a profound impact by the end of 2019.
AIM: What are some of the AI-based products or services that you rely on the most?
HD: In my opinion, the so called “AI” based products are still in their infancy, especially chatbots and virtual assistants. Having said that, I use Ganglia (Cricket Assistant by MACHAAO) everyday for my cricketing needs. I also like to use Amy, the AI personal assistant for calendar scheduling assistant. Of course, then there is Alexa!
AIM: What is Machaao doing in the AI space right now? Please highlight some use cases.
HD: We are currently working on enhancing our machine learning personalisation algorithms to understand and engage with the user to make Ganglia a live AI-friend with a personality that you can talk, play and earn with. By the end of 2019, we hope to have an AI-version of Ganglia that would be able to understand and express various emotions. This essentially means that it would be able to have long detailed match conversations about a particular cricket match.
We are looking at AI to help us enhance personalised content consumption and monetise on sports-related content to offer best user-experience to people for the sports they love.
AIM: What are the areas of life or employment where you would like AI to be more involved
HD: I believe AI will impact every aspect of our lives, including the way we work. It is also likely to impact a lot of industries including customer support, e-commerce, transportation, entertainment among others. However, I feel healthcare has been on the AI bandwagon for a long time now.
AIM: Will AI take away the creative thinking and downgrade the intellectual quotient of humans?
HD: On the contrary, I feel the opposite. AI could be a great tool to increase human productivity and scope for thinking. The best example I could give is of the prediction-game we have been working on. Fans spend infinite amount of time watching or following their favorite sport. This, in probability cannot be classified as productive, but what if they were getting paid for spending their time! Enhanced AI are helping brands facilitate trust and establish a deeper personal relationship with their customers which should lead to a plethora of micro transaction economies.
AIM: Many experts have warned against AI taking over every aspect of our lives. How true is their fear, according to you?
HD: I don’t think the experts are totally unwarranted in their fear, caution is the name of the game. But as the case with any new technology is, they tend to cut some jobs in the process of creating new better ones. Remember, our favourite “iceman” being replaced by “the refrigerator” in early 1900s. Evolution is a gradual process and I think above example should eliminate some fears.
AIM: With voice-based assistants, facial recognition and other AI-based tech being rampant, how will the user privacy and data be jeopardised?
HD: User privacy is and will always remain an issue. That being said, I think legislations that have been put forth by multiple nations around the world would help maintain some sanity as we move forward.
AIM: How has been the adoption of AI in Indian scenario?
HD: What gives us a lot of hope is the fact that we have seen the adoption of a lot of good chatbots. For instance, Ganglia is currently serving 1M+ users from across the globe with a majority of users from India. There are other major bots such as NearGroup with 10M+ users, Mitsuku, Ruuh, the desi virtual friend from Microsoft and others, that are gaining major traction in the Indian market. As the developer community works towards enhancing AI quotient, you can expect enhanced virtual assistants and AI-based products in the coming months.
Also, with the advent of new consumer channels such as Alexa, Google Home, Whatsapp, Google Assistant, etc., we see AI gaining mass popularity.
AIM: What are some of the challenges that come on the way while adopting AI?
HD: I think the biggest challenge would be user “trust” as we move forward into the second generation of AI. There are many examples of trust and goodwill between brands and it’s customer, Apple being a prime example. This trust has been established over decades and the same applies for any business or consumer-driven AI agents.
AIM: Will AI be capable of exhibiting emotions and get a conscience?
HD: Yes, I think AI would be able to exhibit emotions and get a conscience. Many companies around the world including India are in the middle of prototyping research, with sophisticated research making a lot of headway. Having said that, the usefulness or harmfulness of a particular AI system is totally dependent on the the intended use of the AI agent.
Try deep learning using MATLAB