AI adoption for SMEs does not always need to be costly: Josephine Teo

SINGAPORE – Amid worries among smaller firms that artificial intelligence (AI) is too costly to adopt, Minister for Digital Development and Information Josephine Teo pointed out that not all use cases require expensive high-performance chips, and there are many models of support for companies.

She was speaking at a plenary session at the Bloomberg New Economy Forum held at Capella Singapore on Nov 19. Titled The New Digital Divide: AI’s Haves And Have-Nots, the session discussed the emerging global divide between those with AI capabilities and those without.

Recounting how a neighbourhood restaurant used an AI assistant app to optimise its promotions, Mrs Teo said: “Not every single­ small and medium enterprise is going to have to need its own access to graphics processing units (GPUs). In this case, it’s just an app that was made available to this little eatery.”

GPUs are costly specialised processing chips that supply much of the computing power needed for advanced AI.

Mrs Teo said Singapore has different methods of helping small and medium-sized enterprises (SMEs) manage the costs of adopting AI.

“Some will come through tools that are made available by different platforms. Some will be through efforts that are targeted at sectors,” she said, highlighting the example of a predictive maintenance tool for precision manufacturers that was developed for the industry by a sectoral AI Centre of Excellence.

Google DeepMind chief operating officer Lila Ibrahim said the research division is working with its colleagues companywide to try and integrate AI tools into existing programmes in order to expand access.

Another key issue raised during the discussion is how to bridge the gap between countries with varying abilities to adopt AI.

Bloomberg Television anchor and editor-at-large Haslinda Amin, who was moderating the plenary session, pointed out that a United Nations report found that over 100 countries, mostly from the Global South, are falling behind in terms of AI.

To tackle that, Ms Ibrahim said Google DeepMind keeps accessibility in mind while developing its models.

She highlighted the unit’s new AI research lab in Singapore, which will spur existing efforts in improving its AI model’s capabilities in understanding South-east Asian languages and cultural nuances.

“AI is not a technology where we have national borders. We launch a model, and it’s available worldwide. So we have to think about everything from access to literacy. It’s actually making the models easy to interact with, things like making it multimodal from the beginning, like what we’ve done with Gemini,” said Ms Ibrahim, referring to the AI chatbot and virtual assistant developed by Google.

Mrs Teo pointed out that countries have agencies to reshape things to be more ready for an AI future by investing in infrastructure. Organisations such as the World Bank, International Monetary Fund and United Nations Development Programme are stepping up to help nations in need.

The area where countries should invest most in is developing capabilities, said Mrs Teo. These include different capabilities like entrepreneurial and engineering skills, as well as for different groups such as the workforce, enterprises and academia.

“There is a whole range of things that need to be built up in order for a country to enable its citizens to benefit from AI and have the best chance of moving ahead,” she added.

“Will you always get to what the next country is able to achieve? Maybe yes, maybe no. But that shouldn’t hold us back from trying.”