How To Crack AI: Spatial Computing To Cyber Systems, Check Deloitte Top Tech Trends List

Today’s artificial intelligence technologies are all pervasive. AI is becoming a fundamental component of enterprise IT and an essential component in the development and provision of new products and services, from the server room to the board room. However, before organisations can properly utilise AI, they must thoroughly align strategy, people, architecture, and—above all—data.

Deloitte’s 16th annual Tech Trends report throws light on the following technological trends that will move from being sensational to foundational over the next 18-24 months in an attempt to crack AI. 

Taking the visualisation of ideas and objects off the 2D screen and adding voice and gesture to the ways people interact with machines is a surging interest. Spatial computing is becoming more popular because of its capacity to involve people and contextualise data. With AI, spatial computing is evolving from a practical training tool that helps a particular group of employees to a corporate profit centre that can provide real-time, sophisticated data analysis and automated changes.

Turning to hyperscalers for large language models instead of building from scratch has helped many enterprises accelerate their AI adoption. But some organisations are turning to smaller, purpose-built models because of security, energy use, agent-to-agent communication, and other specific needs, the report noted. Working together, several small models can conduct simulations, produce multimodal outputs, handle discrete tasks, and provide users with various virtual assistants.

In order to enhance user capabilities and future-proof the digital infrastructure, manufacturers are working on a new generation of processors that integrate AI models into PCs and edge devices for localised, offline use. Additionally, onboard AI can strengthen Internet of Things in domains like robotics and medical devices. Companies that shifted from hardware at the core may end up making fresh investments in hardware at the edge, the report said.

Quantum computing’s decryption power could make current cybersecurity practices a liability, the report said. From identities to finances to communications—anything that can be entrusted to computers—is at stake. In order to ensure protection, it will be up to organisations to reimagine their cyber mindset and work on a solution in advance.

Integrating AI into core enterprise architecture means change in systems and processes. Enterprises continue to rely on—and invest in—legacy custom systems, ERP and customised cloud solutions. AI can and will be embedded into those systems but might also lead to a recast of the “core” content, data, and transactions, especially as AI trains on data from across the organisation.