Latest news with #NextGenerationAIDevelopmentPlan

The Wire

Things that China Is Doing With AI That You Need to Know

Menu हिंदी తెలుగు اردو Home Politics Economy World Security Law Science Society Culture Editor's Pick Opinion Support independent journalism. Donate Now World Things that China Is Doing With AI That You Need to Know Sundeep Narwani 4 minutes ago In contrast, India's foray into synthetic data remains in its infancy. 'China's comprehensive AI policy framework, epitomised by the 2025 Next Generation AI Development Plan, ensures coordination among academia, industry, and infrastructure.' Photo: Elise Racine & The Bigger Picture / / Real journalism holds power accountable Since 2015, The Wire has done just that. But we can continue only with your support. Contribute now As artificial intelligence (AI) reshapes global economies, China has established a commanding lead, leaving India striving to bridge the gap. Despite India's notable advancements, structural challenges and execution delays suggest that China will maintain its dominance in the AI sector for the foreseeable future. The race to harness artificial intelligence (AI) is intensifying globally, with China and India emerging as significant players. However, a closer examination reveals a growing disparity between the two nations, with China establishing a formidable lead that India may find challenging to bridge in the near future. China's Synthetic Data Supremacy China's AI advancement is notably propelled by its prowess in synthetic data generation. Chinese firms have developed sophisticated generative adversarial networks (GANs) capable of producing training datasets that closely mirror real-world data. This capability is particularly beneficial in sectors like healthcare, where privacy concerns limit access to real patient data. For instance, Chinese medical AI platforms can simulate rare disease progression across thousands of synthetic patient profiles, facilitating robust diagnostic model training without compromising patient confidentiality. This technological edge is underpinned by substantial investment. In 2024, China's synthetic data market reached $1.2 billion, with a significant majority of AI startups integrating synthetic datasets into their development processes. Tools like Alibaba's SynthMaker and Baidu's DeepFabric have become instrumental, automating a large portion of synthetic data workflows for commercial applications. India's Nascent Endeavours In contrast, India's foray into synthetic data remains in its infancy. While startups such as IndikaAI are making strides, their models currently achieve lower accuracy in replicating complex data patterns compared to their Chinese counterparts. The Indian market, projected to reach $158 million by 2030, lacks indigenous platforms tailored to the country's diverse linguistic and demographic landscape. Consequently, a vast majority of Indian AI firms rely on Western tools that may not be optimized for local nuances. Challenges also persist in generating synthetic agricultural datasets essential for crop-yield prediction models. Indian efforts often fall short in capturing microclimate variations across the nation's numerous agro-ecological zones, leading to higher prediction errors relative to Chinese models. Infrastructure: A tale of two strategies China's strategic establishment of AI data hubs in tier-III cities like Guiyang and Lanzhou has proven effective. These centres benefit from lower operational costs and municipal incentives, enabling the processing of vast amounts of training data annually. Such decentralization facilitates continuous data refinement, exemplified by Wuhan's hub processing hundreds of millions of facial recognition samples monthly. India's initiatives in tier-II and III cities, while promising, face hurdles. New data centers in cities like Coimbatore and Jaipur operate below capacity due to inconsistent power supply and limited demand from major cloud service providers. Additionally, state subsidies cover a smaller fraction of infrastructure costs compared to China's support, compelling Indian startups to allocate more capital to data operations. The limited reach of edge computing networks further hampers the deployment of latency-sensitive AI applications across the country. Policy execution: divergent paths China's comprehensive AI policy framework, epitomised by the 2025 Next Generation AI Development Plan, ensures coordination among academia, industry, and infrastructure. This alignment has resulted in a significant proportion of AI patents originating from state-industry partnerships and expedited deployment of technologies like autonomous vehicles through streamlined regulatory processes. India's National AI Strategy, though ambitious, grapples with execution challenges. Initiatives such as the IndiaAI mission have yet to fully deliver on promises like subsidised GPU access for startups. State-level projects encounter delays due to bureaucratic hurdles, and patent approval times for AI innovations lag behind China's, dampening commercial incentives. Linguistic diversity: A complex challenge India's rich tapestry of languages and dialects presents unique challenges for AI training. Current natural language processing (NLP) models cover a limited portion of linguistic variations, in contrast to China's broader coverage across Mandarin dialects. Efforts like Bhashini's crowdsourcing platform face issues with data quality, and initiatives such as Karya's ethical crowdsourcing model operate on a relatively small scale. Furthermore, Indian AI startups often depend on global language models for regional language support, leading to increased costs and dependencies. The road ahead Bridging the AI divide requires India to address several critical areas in the next few months: Synthetic Data Development: Accelerating the creation of robust GAN frameworks tailored to India's diverse data needs. Infrastructure Enhancement: Expanding and optimising data centres, particularly in tier-II and III cities, to ensure consistent power supply and attract major cloud service providers. Policy Implementation: Streamlining bureaucratic processes to expedite the execution of AI strategies and reduce patent approval times. Linguistic Inclusion: Investing in NLP models that encompass the full spectrum of India's linguistic diversity, supported by large-scale, high-quality data annotation efforts. While India's AI ambitions are frequently spotlighted in policy papers and public relations campaigns, the tangible progress remains limited. Without addressing these foundational challenges, the gap between China and India in the AI domain is poised to widen further in the coming years. Make a contribution to Independent Journalism Related News Why Artificial Superintelligence Might Be Humanity's Best Hope India and China: Two Contrasting Models of Dealing With Trump's US India Rejects as China Assigns Names to Places in Arunachal Pradesh Again The China Factor Can't Be Ignored in the India-Pakistan Conflict Over Kashmir China's Latest White Paper is on National Security in the New Era How Indian Media Sabotaged its Own War Efforts As US Steps Back From Tariff War With China, What You Need to Know Trump Terms US-China Tariff Talks in Geneva a 'Very Good Meeting', Says Negotiated 'Total Reset' Chalk Dust and Algorithm: Being a Teacher in the Age of Google and AI About Us Contact Us Support Us © Copyright. All Rights Reserved.