What Singapore businesses need to get ahead in the AI race

IT'S no secret that artificial intelligence (AI) has the potential to transform the way we live, work and play. In fact, it's already happening all around us in Singapore, from chatbots being trialled in managing hospital appointments to robotic lawn mowers cutting grass in the Botanic Gardens.

With the global AI market expected to generate upwards of US$35,870 million in revenue by 2025, it's no surprise that Singapore is taking every action to ensure a strong position in the AI race.

According to a recent report by Accenture, AI could add up to US$215 billion in gross value across 11 industries in Singapore by 2035, reinvigorating the city-state's economy via increased efficiency, productivity and investment.

This is a fact that has not been overlooked by the government, which identifies AI as one of the four core technologies essential to the country's push towards being "digitally ready" as part of its broader Smart Nation Initiative.

AI can offer businesses a wealth of operational benefits, increasing efficiency, ability to scale and to make better business decisions, as well as insight into consumer behaviour and experience, faster innovation and better products and solutions.

In fact, Singapore is well-positioned to occupy a leading role in the AI race, owing largely to its pro-business environment, excellent infrastructure, connectivity to major Asian economies, as well as the availability of investment and well-developed technologies - which makes a compelling case for businesses and top talent thinking to make Singapore their home.

Not only this, but Singapore also recently became the only Asian nation to announce the development of a Model AI Governance Framework.

This is an accountability-based framework to help chart the language and frame the discussions around harnessing AI in a responsible way, translating ethical principles into practical measures for AI deployment at scale, while helping the country to continue investing and growing its AI capabilities while remaining globally competitive.

Despite the AI buzz however, most Singapore businesses continue to face significant challenges when approaching AI initiatives.

In fact, if one looks at it from a macro-level, only a handful of companies in the world have been successful with AI. We call it the "1 per cent problem", as the other 99 per cent are still struggling.

A recent study carried out by IDC, points out that as AI adoption rates in Asean soar - having almost doubled since last year, Singapore does not stand out as the trailblazer we might expect, coming in behind both Indonesia and Thailand for number of businesses successfully implementing AI initiatives.

Truth is, we are seeing enterprises facing the same issues the world over. Last year, we commissioned a survey through IDG's CIO Research Services, which asked 200 IT executives at larger companies (more than 1,000 employees) across the US and Europe about their major challenges when it came to pursuing AI initiatives.

The study revealed that a huge 96 per cent of respondents said "data silos" was their No 1 challenge - which, put simply, represents businesses' struggle to harness the vast streams of reliable data coming from various areas of their business in order to retrieve meaningful actionable insight, which is the cornerstone of AI.

The second problem is the disconnect between data scientists and engineers - with 80 per cent of respondents citing "collaboration" as a major blocker to AI success.

According to the findings, companies take an average of seven months to bring AI initiatives into production, with most using an average of seven different machine and deep-learning frameworks and tools - adding to the overall complexity of the process.

But what does this mean in practice and what do Singapore businesses need to do to overcome these challenges? Put simply, it's all about unification.

The first action point is about unifying data. In any organisation, enterprise data is collected and stored across hundreds of systems that are not necessarily joined-up or AI-enabled, which means that an enormous amount of time is needed in order to combine, clean and verify the data before it can be put to use.

This creates a major bottleneck in the AI journey - and often, by the time the data has gone through this process, data scientists find that the results are not good enough and have to go back to square one.

This can be an expensive and frustrating process, which can only be solved by connecting the dots between big data and machine learning.

This is where Unified Analytics comes in. A new category of solution, Unified Analytics, works to unify data processing with AI technologies, essentially joining up the dots between all the data and cutting through the noise - to allow enterprise organisations to retrieve meaningful actionable information from massive amounts of data.

The organisations that succeed in unifying their domain data at scale and unifying that data with the best AI technologies will ultimately be the ones that succeed with AI.

Unifying teams

The second action point for Singapore enterprises is about unifying teams.

Put simply, much like data, the teams assembled to deal with the data are often siloed from each other - working in isolation, with data engineers dealing with large-scale data processing and production deployment and data scientists dealing with AI - exploring data, training and validating models.

This organisational separation creates friction and slows projects down, becoming an impediment to the highly iterative nature of AI projects.

Keeping these two factors in mind, the future of AI may be bright, with the potential to change Singapore and the wider world, but we won't get there until closer collaborations happen between developers, data scientists and engineers - which is the lynchpin in optimising AI workflow, increasing productivity and ultimately, innovation.

It's those businesses that can successfully tackle this problem and leverage the best available technologies and frameworks to extract valuable insights from data, who will succeed in being the real AI innovators.