Qualcomm is championing a significant shift in how consumers will interact with technology, moving beyond the current app-centric model towards an era defined by agentic AI. Executive Vice President and Chief Marketing Officer Don McGuire outlined this vision, emphasizing on-device AI, secure commerce, and the transformative role of 6G connectivity, particularly highlighting India's potential in this evolution.
The Shift to Agentic AI
McGuire describes this transition as moving from a generative AI age to an “agentic AI age,” where artificial intelligence not only answers queries but actively performs tasks on behalf of users. Qualcomm's strategy involves running more AI workloads directly on devices, enabling AI agents to carry out complex actions and eventually handle commercial transactions.
This vision extends beyond smartphones, encompassing a broad “ecosystem of you” that includes laptops, cars, smart glasses, watches, and earbuds, all powered by Snapdragon silicon. The core challenge for widespread adoption, McGuire notes, will be ensuring these AI agents make daily tasks demonstrably easier, faster, and more efficient than existing methods.
Securing Commercial Transactions
A crucial aspect of agentic AI is enabling secure commercial transactions. While AI can demonstrate booking flights or hotels, allowing agents to complete actual purchases requires robust authentication and payment security. McGuire acknowledges that the “last mile” of commerce—trusting an AI with payment details—is a significant hurdle.
He points to collaborations with financial institutions like Mastercard as a pathway to building this trust, drawing parallels to how consumers adopted ATMs or mobile payment systems like Google Wallet and Apple Pay. India, with its advanced digital payment infrastructure and widespread use of UPI, is seen as a key market where trusted financial brands could lead the way in integrating AI-driven commerce, provided strong user confirmation and authentication protocols are in place.
Why On-Device AI Matters
Latency is a critical factor for agentic AI adoption. If an AI agent takes longer to complete a task than a user would manually, its utility diminishes. McGuire argues that running appropriately sized AI models directly on devices, rather than constantly sending requests to the cloud, dramatically reduces latency.
This on-device processing capability for models up to 20-40 billion parameters could fundamentally alter the role of traditional apps and operating systems. McGuire envisions app stores evolving into directories of interconnected agents, and smart glasses offering entirely new, voice-based interfaces that interpret visual and audio data without a legacy operating system.
6G and AI-Native Connectivity
Looking ahead, Qualcomm recognizes that future AI experiences, especially those involving continuous data streams from devices like smart glasses, will demand significantly greater upload capacity than 5G can provide. McGuire explains that 6G will be crucial for enabling these “AI-native” networks, supporting massive, constant uploads required for always-on AI interactions.
He highlights India's progress in 5G infrastructure as a strong foundation, positioning the country to be as ready as any other for the next generation of connectivity if global standardization efforts hold true for 6G.
Open-Weight Models and Practical AI
McGuire also discussed the economic implications of AI models, noting the rise of open-weight models. While not necessarily surpassing frontier models in raw capability, these models drive efficiencies by enabling deployment for specific workloads without the prohibitive costs associated with relying solely on the most powerful, expensive systems. Qualcomm is also investing in its own on-premise data center infrastructure to reduce reliance on external AI providers.
He observes that China is particularly advanced in applying open-weight models to consumer products and agentic experiences, citing examples within the automotive industry where AI agents are deeply integrated for intuitive user interaction.
Making AI Useful, Not Just Explaining It
Ultimately, for AI to become mainstream, McGuire believes the focus must shift from explaining the technology to demonstrating its practical value. He suggests that AI has long worked in the background, enhancing experiences like photography, without users necessarily knowing it was AI. This “show versus tell” approach is also key to building trust. Rather than simply stating AI is trustworthy, companies must prove it through reliable and beneficial products, especially given the relative newness of many AI entities in the consumer space.