AI has only just begun its journey
2 JAN, 2026
By Anjali Bastianpillai, Thematic Equity Specialist at Pictet Asset Management
There are 20 billion connected devices worldwide. This number is expected to double over the next five years. This trend is creating enormous demand for software components and data centres. Automation and robotics are also benefiting from growing labour shortages in Western economies and from the reshoring of manufacturing. These forces may increase levels of fixed capital investment. Cybersecurity remains the top priority in information technology spending. Taken together technology as a share of GDP could double over the next decade according to Microsoft CEO Satya Nadella.
Artificial intelligence and automation have the potential to boost productivity reduce costs and help address challenges linked to the global labour shortage. Yet they are still at an early stage. So far only one third of computing workloads which are expanding due to AI have moved to the cloud. AI has only just begun its journey. It is becoming a structural technology that will affect every sector of the economy. The key is to make the right choices along the technology adoption curve.
The first requirement is infrastructure. Most AI investment is flowing into semiconductors servers data centres networks and storage. This market could grow at a 28 percent compound annual rate and reach 460.5 billion dollars by 2033. Four major hyperscale platform providers Alphabet Amazon Meta and Microsoft account for around 78 percent of global cloud capacity. Together they have committed 390 billion dollars in capital expenditure in 2025.
Smartphones computers vehicles and industrial equipment all rely on semiconductors. This supports ongoing demand for conventional central processing units or CPUs which still perform most computing tasks. Demand for memory chips also remains strong. Manufacturers are competing to make chips smaller more powerful and more energy efficient. This is especially important because AI requires vast amounts of data processing capacity storage memory and faster connections. Generative AI models rely on high speed graphics processing units. Hyperscalers are developing custom chips to improve performance and costs across pre training training and inference. The growing adoption of AI applications also supports semiconductor producers. The next cycle includes AI integration into smartphones and personal computers. Memory is becoming an increasingly important part of the value chain.
Overall revenues in the global semiconductor market could reach one trillion dollars by 2030. As most production is outsourced there is significant growth potential in advanced semiconductor design and manufacturing as well as in assembly testing machinery and design software. Semiconductor demand is also supported by electrification and autonomous driving.
Generative AI has already changed how robots learn and interact with their environment. A growing market is emerging for agentic AI applications that operate without human intervention. AI is also being embedded into physical systems. These technologies will shape the future of automation and robotics. Examples include autonomous vehicles and humanoid robots. Productivity gains and cost savings from these advances could generate 920 billion dollars in value for companies in the S and P 500 according to Morgan Stanley.
Data centres and AI infrastructure providers also require strong physical and digital security. Hackers infiltrate systems every 39 seconds. An estimated 3.8 million records are stolen each day. AI driven cyberattacks are becoming more sophisticated and harder to detect. At the same time data is often stored across disconnected systems. Lakehouse architecture helps bring data and processed content together on unified platforms. Global cybersecurity spending is expected to grow by nearly 14 percent per year according to Gartner. Companies offering advanced security solutions are likely to be among the main beneficiaries.
