55 MARTINDALE CENTER FOR THE STUDY OF PRIVATE ENTERPRISE ucts as physical goods, such as machines, buildings, and vehicles (Den Hertog et al., 1997). For instance, a tangible product would be a manufactured car, whereas an intangible product would be the blueprint of the car’s design. In the former case, a physical product is sold versus in the latter, a concept of a product. Intangibles not only are a means of circumventing external pressures related to the physical nature of manufactured products but also hold significant value. Intangible assets have grown dramatically in importance—comprising 17% of assets of S&P 500 companies in 1975, they now account for over 90% of total assets of these companies (Brown, 2023). Possible contenders A technology industry providing intangible products and services is the most viable contender for Taiwan’s future. Industries that fit this criterion lie within the umbrella of cloud computing services, the external storage of data and the execution of software in a shared network of distributed servers (Rashid & Chaturvedi, 2019). An example is Amazon Web Services, which provides access to computing, power storage, databases, and other information technology resources synchronized across multiple devices for specific users via purchasing options provided by Amazon Web Services. Cloud computing can be divided into three service sectors: infrastructure, platform, and software. An example of a cloud computing service that contains these sectors is Microsoft Azure. Infrastructure as a Service has the most fundamental computing resources: features like CPU, memory, and physical servers provided by Microsoft Azure. Platform as a Service includes operating systems that act as the framework for the services, such as the Microsoft SQL Server database. Software as a Service (SaaS) includes the ultimate application for use, such as Microsoft Office 365 (Rashid & Chaturvedi, 2019). SaaS is considered the dominant of the three sectors. The SaaS market generated over US$152B in end-user spending in 2021, trumping Infrastructure as a Service, at US$91B, and Platform as a Service, at US$86B (Slingerland, 2024). As SaaS continues to grow, so does the presence of AI in the sector, because AI allows SaaS companies to gain a competitive edge, including advanced personalization in AI algorithms to analyze vast amounts of user data, task automation to reduce users’ time and streamline business processes, and predictive insights on user behavior for ease of use (Rrucaj, 2023). Among SaaS businesses, 35% currently utilize AI and an additional 42% seek to implement it in the future (Page, 2024). AI interest Global AI software revenue is projected to reach US$118.6B in 2025, up more than 1000% from US$9.5B in 2018. Moreover, AI is predicted to become integrated into almost every new software product and service by 2026 (Page, 2024). Such significant implementation is attributed to not only tech giants but also the bountiful presence of AI start-ups generating such software over the past two decades. In 2011, the total investment in AI start-ups globally was only US$26M across seven start-ups (Soni et al., 2020), which increased exponentially to US$100B in 2024 (Teare, 2025). And it is still growing. In response, there has been a recent interest by Taiwan in expanding into AI development. As President Lai Ching-te recently stated, there is a desire to expand from a “silicon island” into an “AI island” (Sharwood, 2024). However, the AI market continues to be dominated by tech giants, particularly those in the US. As of 2024, the US was responsible for 83% of new AI unicorns (Wright, 2025); in Q1 2025, North America garnered nearly 90% of AI venture capital funding (Hoffman, 2025). Therefore, if the small island is to compete with AI leaders such as Meta, Google (Gemini), OpenAI, and Nvidia, Taiwan must bring something new to the table. Defining AI To evaluate Taiwan’s potential for AI investment, it is important to clarify exactly what AI is. AI is the creation of intelligent systems (McCarthy, 2004). What differentiates AI from natural intelligence is that AI is created by humans and is not of biological or evolutionary origin. AI is not to be confused with machines, which are simply objects capable of doing work (Fetzer, 1990). When analyzing AI, the definition of intelligence is often debated depending on the perspective used, such as philosophy versus psychology perspectives. A generally accepted definition is that intelligence is the ability to learn or understand from experience (Fetzer, 1990). Thus, AI can be thought of as a system created by humans that has the ability to learn. AI has essentially almost infinite applications, yielding its immense potential. Current applications for AI are gaming, robots, weather forecasting, statistical analysis, data mining, knowledge representation, and medical diagnostics, to name just a few. From NASA’s AI robots in space exploration (Borana, 2016) to autonomous vehicles, the applications of AI are rapidly expanding, and so is its demand.
RkJQdWJsaXNoZXIy MTA0OTQ5OA==