Cloud giant AWS has unveiled a jaw dropping S$12 billion investment in Singapore to turbocharge cloud and AI adoption across the city-state over the next four years.
Under a new collaboration called AWS AI Spring 5000, individuals will be groomed annually in AI skills between 2024 and 2026.
AWS has also partnered with the country’s Infocomm Media Development Authority (IMDA) to nurture digital leaders working in bigger local enterprises, to identify potential AI business solutions, navigate technical complexities and deploy their own customised GenAI solutions.
To prepare students to be AI-ready, the cloud giant is tying up with the Institute of Technical Education and polytechnics to integrate machine learning and other AI tools into their curriculum, prepping students to emerge AI-ready for the workforce. The AWS AI Spring programme was unveiled during the 10th AWS ASEAN Summit held yesterday.
With the new investment, AWS’ total planned investment into its existing cloud infrastructure will double to more than S$23 billion, according to the company.
Since it opened its office here 14 years ago, AWS has entrenched itself in the local technology scene, gradually expanding its operations in the city-state. It has also boosted digital expertise here, helping to train 400,000 professionals in cloud skills since 2017.
It has also worked on several projects including Sea-Lion, a large language model (LLM) akin that is trained to understand Southeast Asian languages and culture.
According to a new AWS Economic Impact Study, AWS’ total investment in Singapore is estimated to contribute S$23.7 billion to the Republic’s Gross Domestic Product by 2028, and support an estimated average of 12,300 jobs in local businesses each year.
These are in areas like construction, facility maintenance, engineering and telecommunications, all of which are part of the AWS data centre supply chain in Singapore.
AI driving the recent S$12 billion investment
According to management consultancy Kearney, AI in Southeast Asia has the potential to add as much as US$1 trillion to the regional economy. With the region in its nascent stages of AI adoption, the potential for growth is huge.
AWS vice-president Deepak Singh said that AI’s growth is intertwined with the burgeoning demand for robust data infrastructure which is critical for building mammoth data lakes, which underpin GenAI applications.
He expects inferencing to be the biggest demand generator for cloud services as organisations grow confident of using AI and begin to draw on compute power like GPUs and other cutting edge tools.
While talent will always be an issue, he said AI will boost software engineers and developers’ efficiency, enabling them to build business solutions quickly.
Singh, who oversees next-gen developer experience at AWS, cited the experience of a small team of AWS data centre technicians who built an app to debug errors in the air conditioning systems of the data centres.
“No one cares about this problem, only this small team,” he noted. “They just query the data depository in a chat box, getting a response very quickly.”
“The app has saved them time, making them more efficient, now they can pay attention to other important tasks,” he explained.
“Previously they would have had to wait their turn for the IT team to provide development support or buy a suitable app from the market,” he told Techgoondu in an interview. “Now they can develop it themselves very quickly using AI.”
Today, there are several tools available like Amazon Q Business, a coding assistant to enable software developers in enterprises to generate code quickly.
In AWS, software engineers, marketers and other executives can query internal depositories – in natural language, not code – using a chatbot that is linked to several LLMs. Depending on the query, the chatbot selects the best LLM to use.
Another new class of tools called AI agents are emerging that will automate software developments tasks and improve work efficiency. An AI agent, for instance, can help software professionals undertake tasks like data transformations very quickly.
“Give the agent a natural language prompt, the agent looks at the code base and recommends where changes can be made,” said Singh.
In future, there will be groups of AI agents who can “chain” together several tasks. A software developer writing a new application can use a chain of AI agents to gather end user requirements, undertake research and design before they begin to generate code.
The chain of AI agents – each one able to complete a specific task– could undertake the grunt work of gathering information which the developer can use to write the code.
“With this, software developers can spend the majority of their time writing code, rather than gathering information,” said Singh, “I believe this will happen in the next couple of years”.
Amid the euphoria surrounding AI’s ascent, concerns persist regarding the opacity of AI algorithms. Critics have been alarmed, for example, by the inability to explain how an AI arrives at a result.
According to Singh, AWS now have the tools to foster transparency and accountability and “demystify what’s inside the box, allowing anyone to ask in a chat box, what the program can do and how the code is being written”.
This is a way to make businesses more comfortable with AI and drive adoption, he added.
