AI Trends for Businesses in 2020

It is essential for any business to keep testing new processes and technologies. This continuously cautious approach lets them stay on pace as entire industries change and evolve, but also gives them options they probably would not consider otherwise.

Take Artificial Intelligence or AI as one example. More Singaporean businesses are now applying AI to automate and speed up tasks, or analyse massive datasets within a shorter timeframe. The International Data Corporation (IDC)’s 2018 survey on AI adoption in the Asia Pacific region had Singapore in third place for organisational AI adoption (9.9%). Indonesia ranked first with 24.6%, and Thailand came in second with 17.1%.

Local and regional entrepreneurs and startups also tout AI as integral to their products and services, and with both planned and unexpected applications. As of August 2019, Tracxn puts the number of AI-centric startups in the city-state at 405.

This shift toward AI is both about the economy and work efficiency. While the ultimate goal is to further improve the city-state’s quality of life with advanced and scalable technology, AI adoption is also meant to address economic concerns such as “slowing growth, falling capital investment, soft workforce growth and decelerated productivity.”

In November last year, the Singapore government made a comprehensive national AI roadmap, with specific national goals to hit for the next two, five and 10 years. As for us, we’re looking at what’s in store for business in 2020—and what companies can do for their own operations.

We partnered with technology research and advisory firm Ecosystm, and published a report on the major enterprise AI trends in 2020 using data on their platform. Out of the five AI trends in the report putting robotic process automation (RPA) and analytics at the forefront in the workplace, two are of utmost importance for Singaporean businesses.

First, automation will lead organisations to AI. This trend goes two ways: businesses that have not applied simple automation or RPA processes yet can and will do so in 2020, while those with existing RPA processes would want to incorporate smarter AI.

Among the AI solutions that will have bigger roles are IoT sensor analytics, machine learning, natural language processing or NLP, smart process automation, and image analytics.

The other business-centric trend is that AI will be embedded in most business applications. Instead of having AI as a separate process, self-learning intelligent applications will be more widely available in 2020. Business solutions already in use will have the capability to parse datasets they contain, as opposed to extracting data and requiring human intervention or an external platform to analyse them first.

Our report adds that “standard ERP, CRM, SCM, knowledge management solution[s] and other business applications will have embedded intelligence.” Having AI built in won’t just make work progress faster, as is the goal of most organisations in implementing the technology in their workflows. It would also decrease the required human labour needed or the number of analytical tools and platforms used, leading to lower costs alongside increased efficiency.

These trends are already in play for Singapore’s vital industries, making the country well-positioned for further AI adoption this year.

Business processes and operations management would be one of the first to leverage AI technology, either in-house or through partnerships. Portcast is one Singapore-based example. This startup employs predictive analytics with real-time signals so maritime logistics companies can better forecast cargo flows. It also helps them account for external risks, such as inclement weather, fluctuating fuel prices, and port delays. With AI, Portcast aims to assist logistics firms in improving their overall efficiency, use of assets, and pricing.

The financial sector is seeing a similar influx of new players. AI is big with digital banks, with alternative lenders touting fast response times and improved user experience, as well as analytics and compliance startups targeting traditional institutions (banks and investment firms) for upgrades. Among the most-used AI applications are facial recognition, optical character recognition or OCR, data mining, and fraud detection.

Quantiply’s Sensemaker AI software combats financial fraud through quick analysis of customer transactions and behaviors and promises to lower false positives by 50%. On the other hand, Taiger is on the automation side of AI in finance, with its solutions applied in banking and insurance firms plus government agencies for both back-end and front-office concerns.

Urban living and transport are also priorities here, and they intersect with other industry applications for AI. The former covers not just automation and communication between the government and the general public. It includes an irregular-movement alert system for the elderly, the creation of smart towns and 3D models, and area surveys via drone for health issues such as dengue.

Meanwhile, AI adoption for transport and mobility will include autonomous and on-demand vehicles (and their controlled testing), and contactless payments.

AI in HealthTech and MedTech will also change the way patients are examined, diagnosed, treated, and managed for both unique cases and chronic diseases. And even as Singapore places third globally in using AI for diagnostics, it wants to do more, preferably as early as 2022 (e.g., SELENA+ for diabetes retinopathy).

There is a growing number of startups working in this space. For example, See-Mode uses computer vision and computational fluid dynamics to predict strokes and vascular diseases. Its first product, Augmented Vascular Analysis or AVA, was approved by the Health Sciences Authority (HSA) as a Class B medical device in December 2019. Biofourmis also personalises patient intervention and management through mobile AI and wearable biosensors.

One of the first (and most repeated) concerns regarding AI in business is that the heavier reliance on machine learning and automation will render existing jobs obsolete.

Studies like research firm Oxford Economics’ 2019 look at global manufacturing jobs being displaced by robots by 2030 aren’t encouraging at first glance. 20 million jobs will be lost, and The Straits Times added that 1.7 million of these have already been replaced by robots since 2000.

However, Singapore is in a good position to adapt to wide-scale changes caused by AI. Local manufacturing and finance are seen as able to evolve with these demands, thanks to the national roadmap, the upcoming AISG AI Certification programme, and plans to retrain employees and build up AI skill sets.

And according to Stanford University’s AI Index 2019 Annual Report, Singapore ranks first in AI hiring, with its hiring rate rising threefold compared to the 2015-2016 average. At least here in the Lion City, the question of job loss has been mitigated.

But several long-standing issues remain. Foremost are the so-called black-box problem (or the lack of transparency and public understanding of how machines actually parse data and make data-driven decisions), bias in decision-making because of limited or ‘dirty’ datasets, and ethical issues for both humanity at large and future technological evolution.

Data privacy, storage, and control issues are also strongly associated with AI, especially with frequent news headlines of database breaches and fraudulent activity.

Even as companies access more consumer data, consumers don’t trust machines to decide for them, or in relation to them. Humans are still the best decision-makers, who can (in ideal situations) go through data with the proper context and comprehension.

Businesses must then emphasise ‘explainable’ AI and access clean datasets for training. The more their target markets know why, how, where, and when their personal data are used—and the more they see they won’t be discriminated against or refused outright for financial, medical, and government services—the better their reception will be towards AI.

In turn, investors need to look for more than a business’ ability to scale through funding stages—and its long-term prospect of making money. Financial backing shows trust in a company, which can falter when the consumers themselves don’t trust that company. Additionally, any misuse of private data—not to mention the resulting bad publicity—can impact investors’ other ventures or their ability to credibly support new ones.

As for ethical issues, Singapore is again a step ahead. Back in November 2018, the Monetary Authority of Singapore (MAS) issued its principles to promote fairness, ethics, accountability and transparency (FEAT) for AI use in the financial sector. The FEAT Principles are meant to provide guidelines for internal governance, and to encourage trust in AI for financial services and implementation.

And in January 2019, the national government introduced a model framework at the World Economic Forum. This framework tackles how AI can be used by private organisations, and how they could respond to any related issues that arise. As it is a ‘living document’, it is subject to change with appropriate feedback and thorough review.

Singaporean businesses can choose to implement specific AI trends starting in 2020. But they will also deal with issues beyond providing life-changing products and services. There is a critical need to balance AI implementation and benefits with a human-centric and ethical view, and the same care for consumer data as in acquired investments.

Similarly, as Singapore steadily marches toward national AI adoption, its proponents must bridge the massive gap between public awareness and AI complexity, and automation versus job creation or adaptation. Machines may be ready to do our work, but we still need to work on them to get it right.

At SGInnovate, we believe in the revolution of AI in the near future and it continues to be one of our main investment areas.

Keep up with our articles on the applications and implications of AI here.