This article first appeared in Digital Edge, The Edge Malaysia Weekly on April 14, 2025 - April 20, 2025

Let’s start with an April Fool’s fable: A group of lifelong friends debated where to meet for their reunion dinner. In their 40s, they chose the Bay Hotel for its attractive wait staff. A decade later, at 50, they returned for the excellent food and wine. At 60, the hotel’s quiet ambience and ocean view won them over. At 70, they agreed to meet there because the restaurant was wheelchair accessible. Finally, when they turned 80, they debated and settled on the Bay Hotel — because none of them could remember ever having been there before.

If that anecdote made you laugh, these statistics should make you blink: information and communications technology (ICT) spending across the Asia-Pacific region is projected to exceed US$1.66 trillion (RM7.36 trillion) by 2028, up from about US$1.4 trillion in 2025, according to recent estimates from International Data Corp (IDC). This represents a compound annual growth rate (CAGR) of 5.8% and highlights a shift towards investments that deliver measurable return on investment (ROI) and enhance resilience against future uncertainties.

The top 10 fastest-growing industries account for nearly 80% of the region’s total ICT expenditure. Companies are accelerating their digital transformation efforts, embracing AI and strengthening cybersecurity to enhance tech resilience. Despite challenges such as trade restrictions, geopolitical tensions and cybersecurity risks, businesses remain focused on improving efficiency, customer experience and competitiveness in an increasingly dynamic digital landscape.

“Businesses in this region are entering a new phase of tech adoption,” says Mario Allen Clement, an IDC associate research manager. “The emphasis is no longer solely on acquiring the latest technologies but on strategically deploying solutions that address specific business challenges and drive measurable outcomes. There is a move towards pragmatic investments to enhance productivity and improve customer experience.”

The uptrend in investments is across all business sizes. Large enterprises are prioritising innovation and operational efficiency. Mid-sized firms are investing in data analytics, artificial intelligence (AI) and machine learning to scale operations and enhance decision-making. Small firms are focusing on cost-effective productivity tools.

The next new phase in town is agentic AI. That refers to AI systems that can operate autonomously, make decisions and take actions to achieve specific goals without constant human intervention. These systems exhibit characteristics such as adaptability, goal-driven behaviour and can interact with their environment dynamically.

A good example of agentic AI is a self-driving car. It perceives its surroundings using sensors and cameras, processes real-time data, makes decisions about speed, lane changes and obstacle avoidance, and autonomously navigates to its destination — all without direct human control. The AI continuously learns from its experiences to improve driving efficiency and safety.

Should you agonise about agentic AI? Yes, because this new tech is poised to revolutionise the way service interactions are conducted. By 2029, agentic AI will autonomously resolve 80% of common customer service issues without human intervention, leading to a 30% drop in operational costs, says research house Gartner.

“Agentic AI has emerged as a game changer for customer service, paving the way for autonomous and low-effort customer experiences,” notes Daniel O’Sullivan, a Gartner senior director and analyst. “Unlike traditional Gen AI tools that simply assist users with information, agentic AI will proactively resolve service requests on behalf of customers, marking a new era in customer engagement. Organisations will need to rethink their approach to managing inbound service interactions and prepare for a future where AI-driven requests become the norm. In this future, automation will need to become the dominant strategy for all service teams.”

So far, so good. But what about April Fool fables? Here are six AI trends in alphabetical order on the “crazy” side that push the boundaries of what AI can do in unexpected, wild, unsettling ways:

● Art: AI is increasingly capable of creating art, music and even literature that rival human creators. Platforms like DALL-·E (for art) and OpenAI’s MuseNet (for music) can generate realistic paintings and compositions based on a simple prompt, raising questions about originality, creativity and whether machines could one day replace human artists. The blurring of lines between human-made and AI-made content is already reshaping creative sectors, not always in ways that humans find acceptable or desirable.

● Bombs: One of the most disturbing and controversial AI trends is the development of autonomous weapons. These systems can identify and engage targets without human input, such as drones or robots making battlefield decisions on their own. Military-grade AI-controlled weapons could lead to devastating consequences if left unchecked, raising ethical and safety concerns about AI use in warfare.

● Clones: AI can now create hyper-realistic deepfake avatars and synthetic voices, even allowing people to “live on” digitally after death. Companies like Replika, Eternime and DeepBrain AI offer AI-generated clones that mimic deceased loved ones’ speech, personality and mannerisms. Ethical concerns aside, this tech blurs the line between reality and simulation, especially if you can dump your brain to a digital twin.

● Dreams: Some researchers are exploring how AI can simulate dreams or even subjective consciousness. Projects like Google DeepDream can generate surreal, dream-like images. Neuroscientists are investigating whether AI can recreate human-like thought processes or even experience emotions. The idea of an AI “hallucinating” or dreaming is fascinating and unnerving. Is “artificial consciousness” the next phase?

● Extrapolate: Researchers are working on AI systems that could interpret and decode brain activity, essentially allowing AI to “read minds”. Neural interfaces like brain-computer interfaces (BCI) are pushing this boundary by enabling direct communication between the brain and machines. The goal might be to influence thoughts or even implant ideas in people’s minds, leading to a future where privacy, free will and mental autonomy could be manipulated or be at risk.

● Fabricate: AI models are now being used to design, train and optimise other AI systems with minimal human intervention. Google’s AutoML and OpenAI’s reinforcement learning models are early examples. This raises the possibility of AI evolving beyond human-designed constraints — potentially leading to AI ecosystems that improve themselves at an exponential rate and fabricate other AI systems in their image.

The adage “garbage in, garbage out” takes on new relevance in the AI era: “bias in, bias out”, especially when AI training data is flawed. Gartner reports that 63% of organisations are either unsure or lack the proper data management practices for AI. A 2024 survey of 1,200 data management leaders found that those who fail to recognise the differences between AI-specific data needs and traditional data management practices risk undermining their AI initiatives.

Gartner forecasts that by 2026, 60% of AI projects will be abandoned due to inadequate AI-ready data.

“IT leaders can no longer rely solely on conventional data management practices for successful AI integration,” warns Roxane Edjlali, a senior director analyst at Gartner. “Traditional data management is too slow, too structured and too rigid for AI teams. Additionally, it lacks proper documentation and data is often fragmented or siloed across various systems, making it difficult to assess its readiness for AI use.”

How can you fix that? To start with, define what constitutes AI-ready data. “The data must be representative of the use case, of every pattern, error, outlier and unexpected emergence that is needed to train or run the AI model for the specific use,” Edjlali says. “Proving the AI readiness of the data is a process and a practice based on the availability of metadata to align, qualify and govern the data.”

The bottom line: As AI continues to evolve and shape industries, it is crucial for government and industry leaders to guide its development with a focus on augmenting — rather than replacing — human intelligence. While AI can enhance efficiency and decision-making, its success depends on the quality and integrity of the data fed into it. Ensuring that AI projects are built on AI-ready data is essential to prevent bias, inefficiencies and failures. It’s ideal that AI complements human capabilities and fosters innovation, rather than diminishing the role of human intelligence in decision-making processes.

Since we started with one April Fool anecdote, let’s end with another: A group of senior citizens decided to get in shape once they had retired from their day jobs. At 65, they started by walking a mile a day. At 75, they added light stretching and balancing to their daily exercise regimen. At 85, they all agreed that just getting out of a chair without making a noise would be counted as a workout.

Raju Chellam is editor-in-chief of the AI Ethics & Governance Body of Knowledge and chair of Cloud & Data Standards, Singapore

Save by subscribing to us for your print and/or digital copy.

P/S: The Edge is also available on Apple's App Store and Android's Google Play.