Techie Tonic: Is the future of AI computing in the devices we carry?

Thanks to my fellow MANY CXO community members for this interesting discussion: Do you think the next major transformation in computing may not happen in the cloud. It may happen in the devices people carry, wear and use every day? Smartphones, laptops, tablets, vehicles, cameras and industrial equipment are rapidly evolving from conventional computing endpoints into intelligent devices capable of understanding users, processing information and taking action.

Historically, endpoint devices were designed to run applications, whether software on a computer or apps on a smartphone. As computing demands grew, more sophisticated tasks were shifted to cloud-based servers. Artificial intelligence is now redefining this approach by bringing greater intelligence and processing capabilities directly to devices.

A growing number of devices are being equipped with dedicated AI processors, commonly known as neural processing units, or NPUs. These chips allow devices to perform certain AI tasks locally, reducing dependence on the cloud. This can mean faster responses, improved privacy and the ability to use AI services even when internet connectivity is limited.

The shift is particularly visible in personal computers and smartphones. Instead of simply opening applications, users can increasingly interact with their devices through natural language. An AI-enabled computer could summarize a meeting, organize files, search documents, generate content or help automate repetitive tasks. In the future, users may not need to know which application performs a particular task. They may simply tell the device what they want done.

This points to an even bigger change: the rise of the AI agent.

Rather than waiting for instructions at every step, future endpoint devices could understand a user's objectives, gather information from multiple applications and services, and complete tasks on the user's behalf. A smartphone, for example, could potentially coordinate a calendar, email, travel information and navigation to organize a business trip with minimal intervention.

Personalization will be another defining feature. Devices could learn an individual's preferences, routines and working patterns, allowing AI systems to provide increasingly relevant assistance. Much of this information could be processed locally, potentially reducing the need to send sensitive personal data to distant servers.

However, the AI endpoint revolution also brings significant challenges. Security and privacy will become more important as devices gain access to increasingly sensitive information. An AI system that can read emails, access documents and perform actions on behalf of its owner creates a new category of cybersecurity risk.

Organisations will therefore need to rethink endpoint management. Traditional IT security has focused on controlling devices, operating systems and applications. In an AI-driven environment, organizations will also have to control what AI agents can access and what actions they are authorized to perform. Complex isn’t it?

The physical form of computing could change as well. AI-powered glasses, earbuds, vehicles and robots could become important new endpoint categories. Cameras, microphones and other sensors will allow these devices to understand their surroundings, making computing more contextual and less dependent on screens and keyboards.

“The economic implications are equally significant. The technology industry is moving toward a competition involving semiconductor manufacturers, cloud providers, operating-system companies, device makers and AI developers. Success may depend not simply on producing faster hardware, but on integrating chips, software, AI models, security and services into a seamless experience”.

Ultimately, the AI era could redefine what an endpoint device means. The computer of the future may no longer be merely a machine that waits for commands. It could become an intelligent partner that understands context, anticipates needs and performs tasks.

The biggest change, therefore, may not be that devices become more powerful. It may be that they become more useful, personal and autonomous, turning the endpoint from a tool operated by humans into an intelligent computing environment that works alongside them.

Also imagine the change of VDI (Virtual Desktop Infrastructure), it is evolving rapidly in the AI era, moving beyond traditional desktop delivery to become an intelligent, secure digital workspace. AI can help organizations automate IT operations, predict performance issues, optimize computing resources, and enhance cybersecurity, while GPU-enabled VDI supports demanding AI, data science, engineering, and creative workloads. At the same time, AI assistants and autonomous agents can work alongside employees within controlled virtual environments, accessing applications and data securely. As organizations adopt generative and agentic AI, VDI is increasingly positioned as a secure foundation for enabling collaboration between people, applications, and intelligent digital agents.

Stay tuned for more updates…