Over the last few years, emerging technologies like artificial intelligence and machine learning have had significant adoption rate across various stakeholders in India.
From startups to NGOs, all the major players have acknowledged the importance of technology in deriving valuable insight. The government has also taken considerable measures to give requisite impetus for the large scale adoption the technology.
In India’s first-ever National Strategy for AI, NITI Aayog has laid a strong foundation regarding the government’s plan to encourage nation-wide adoption of the technology. Addressing the present challenge of not having the requisite public data set for private as well as public establishments to deploy AI and ML, NITI Aayog’s AI policy stated that it will create a public portal for accessing data set.
Called the National AI Marketplace (NAIM), the three-pronged formal marketplace will be created focusing on data collection and aggregation, data annotation and deployable model. The key concept behind the platform is to ensure a level playing field for all the major AI players and to promote healthy adoption of the technology.
“The marketplace mechanism being proposed here will be aimed at easing the adoption efforts of all participants – private enterprises, PSUs, governments, startups and academia. A common platform which brings together enterprises and AI solution builders will trigger off the initial collaboration towards building AI solutions and adopting them at scale,” the draft policy said.
Though it has been almost a year since the historic announcement, the proposed market place is yet to be designed and it would take the government a few more years to materialise the plan, NAIM is truly a one-of-a-kind initiative which will have larger implications on the AI and ML market. By enabling access to its vast pool of public data which is presently lying dormant, the government can ensure that the key players can make the best use of these data.
In this article, we take a look at some of the salient features of the digital platform and why we think the government should be working towards making the plan a reality before India misses out on the first wave of development in the field of AI and ML
Data marketplace: The marketplace aims to address the issue of the lack of price discovery issue that is presently attached to the current informal data acquisition mechanism. By establishing a formal data market, the government hopes new entrants in the AI model creation or training space will find it easier to raise resources for buying data. “This will be due to VCs having enough information to verify the funding demands and lower risk resulting from a more level playing field and equitable access to new data sources,” it said.
Data annotation market: The purpose of establishing this platform is to reduce the cost attached to the present annotation market. Following the lines of existing players Amazon’s MTurks, Figure-Eight which crowdsources-trained or non-expert anonymous annotators on the internet, the government is mulling a citizen participation platform called myGov.in where seed data for annotation will be made available. In order to achieve this, vast public datasets that are available at organisations such as ISRO (India Space Research Organisation), ICAR (Indian Council of Agricultural Research), All India Radio (AIR) and NIC will be used by the government.
“It is envisaged that these datasets be used to create initial corpora of datasets, and encourage enterprise action in the area,” the policy reads.
Data model marketplace: This marketplace will consist of AI solution developers like startups, enterprises, individual researchers etc and aims to bring together the buyers and suppliers of AI solutions for easy adoption of AI. “It would give visibility to the existing solutions, address information asymmetry and generate awareness amongst the relevant constituents. This AI-as-a-Service model segment would further build upon and synergise with the previous two layers of the marketplace model viz. data and annotation,” states the policy.