AI is creating a new normal for work and play

Source: Dell Technologies.
Clarke.

Artificial intelligence (AI) is going to extend into every aspect of our lives, whether we know it or not. Today, our devices can rely on software using AI and machine learning to optimise power and compute resources based on individual usage patterns, says Jeff Clarke, COO and Vice Chairman, Dell Technologies. Tomorrow, they will do much more.

AI is getting better

“Over the (current) year, these advancements in AI and machine learning will turn our PCs into even smarter and more collaborative companions. They’ll have the ability to optimise power and battery life for our most productive moments – and even become self-sufficient machines that can self-heal and self-advocate for repair – reducing the burden on the user and of course, reducing the number of IT incidents filed,” Clarke predicted.

Source: Ramco Systems.
SivaSubramanian.

“Machine learning-based applications have evolved from email spam and malware filtering to predicting and recommending actions. 2020 will see companies benefitting from machine learning/deep learning’s real-time financial applications such as fraud/anomaly detection, self-auditing systems and suggestions based on past orders/transactions. The consolidation of operations will prevent leakage and mistakes," said Ramesh SivaSubramanian, Head of Ramco Innovation Lab – Singapore, Ramco Systems.

Source: Infor. Rider.

Rick Rider, Senior Director, Product Management, Infor, calls the shift the 'new normal'. He said, “In 2020, AI and machine learning platforms will start to challenge conventional thinking, when it comes to enterprise business processes and expected outcomes. In other words, these systems will redefine our default assumptions about what is 'normal'. This will make business process re-engineering and resource training more efficient.

“When examining supply chain processes, for example, AI platforms have observed that default values – related to expected delivery dates and payment dates – typically are used only 4% of the time. Users almost always plug in their own values. Therefore, AI and machine learning systems will start enabling us to disregard default values, as we understand them today, and act more quickly through trust in our data.”

Improving at personalisation

Charles Ng, VP of Enterprise AI, Appier, said that AI will become more creative at personalisation. "There are many elements to personalisation, such as identifying the right people to reach out to, determining what to say to them and the best time to say it. AI is already doing a great job of this; in fact our own AI-powered platforms are helping many consumer brands do this effectively.

“At this stage, however, the creative aspect of consumer outreach remains something of a bottleneck. It still requires a human to take the insights mentioned above (who, what, when) and manually create the content (images, words, colours) before distributing it.

“In the next few years, we can expect to see this element start to become automated using AI. We will see AI take on more of the personalisation work by generating more creative content - images, stories, narratives, etc.- that go beyond its current core strengths of prediction and detection.”

Michael Weingartner, CTO, SAP Concur, calls the concept 'empathetic AI'. “Customers are individuals with similar needs: to feel important, listened to and respected. As a result, empathetic AI is increasingly applied in advertising, customer service, and to measure how engaged a customer is in their journey. For example, are they attentively focused or just passively scrolling?

“Consumers are already benefitting from this trend, through music streaming services that suggest artists, songs or playlists based on your listening history, or from digital vendors that suggest items of potential interest, based on your past purchases. In 2020, this trend will kick into much higher gear, with more technology companies infusing empathy into their AI.

“Expect to see it more often in the enterprise technology space, as well. As companies use empathetic AI to bring more of the benefits of advanced technology to life, they will instill more trust, create better user experiences, and deliver higher productivity.”

Rider agrees, at least for digital assistants. “Users no longer are satisfied with just telling digital assistants what to do and having them automatically execute certain tasks or basic configurations. 2020 will be the year when these digital assistants, using AI and machine learning, start to understand the context of what users are doing, recommend potential next steps (based on completed actions), identify mistakes and autocorrect inputs, and start to engage with users in dynamic, on-the-fly conversations,” he said.

Vic Sithasanan, Chief Growth Officer & Co-founder, Hyperlab, an Everise company, said branding will evolve because of AI. “As AI powers more and more experiences, the inevitable result is that brands evolve from monolithic entities into the bots that are the primary interface between the business and the customer. Consumers will see bots as brands,” he suggested.

Many of these achievements are only possible with machine learning (ML). AG Lambert, SAP Concur Senior Vice President, Spend, Data and Analytics says that ML will become so ubiquitous that it will “fade into near invisibility, even while making a huge impact”.

“Technology services will increasingly anticipate your needs, whether they be related to expense reports, scheduling, or other processes. Eventually it will do all or a portion of certain tasks, with more accuracy and speed than a human being. SAP Concur is already working on advances in these areas. Your preferred travel itinerary may be suggested and filled out based on previous trips, and automatically combined with your company’s preferred vendor commitments,” he said.

Operationalising AI

Source: JOS Singapore.
Tan.

If the economy is uncertain, AI might help. Andrew Tan, MD at JOS Singapore said, “While it might seem cliché, this continues to be a game changer across industries. One of the main benefits of AI is to help businesses operate in a world of synchronous time, where customers want to be able to schedule medical appointments, or enquire about a product simply by speaking to a chatbot. ”

Several vendors see industry verticals benefitting greatly from AI in 2020. Rider predicts that industry-specific templates will make AI easier to use and deploy in 2020, for example.

“In manufacturing, AI and ML systems will take advantage of templated processes to help enterprises better manage their parts inventories, improve demand forecasting and supply chain efficiency, and improve quality control and time-to-delivery.

“In healthcare, organisations will leverage AI and ML to better integrate data that’s segregated in application silos, exchange information with partners across the care continuum, and better use that data to respond to regulatory and compliance requirements.

“And in retail, companies will use AI and ML to better predict demand patterns and shipment dates, based on defined rules, and improve their short- and long-term planning processes,” he said.

Adrian Jones, Executive VP, Automation Anywhere, said that the public sector will sharpen its focus on supporting AI adoption in the workplace. “We already see that in countries like Singapore, with the recently-announced National AI Strategy which will support key sectors like education, transport, and security. We believe that we will continue to see enquiries on implementing digital workers from public sectors across Asia Pacific and Japan,” said Jones.

Source: ABC Technology. Yang.

AI platforms that can comb unstructured documents for insights will be a boon, added Mason Yang, CEO of ABC Technology. ABC Technology offers such a platform for financial firms. Jones takes the concept further, saying that use cases of robotic process automation (RPA) would eventually move towards easy-to-deploy-and-teach AI models for a large majority of paper documents in the enterprise.

“This means that human workers may never need to process an invoice again. This is due to the onset of rapidly-maturing machine-learning technologies that understand the written word, as well as integrating resulting actions within an RPA platform,” he said.

Asheesh Mehra, Co-Founder and Group CEO, AntWorks, an AI and robotics specialist, says that there is a lack of awareness over the types of solutions that are available to process unstructured data. "While organisations and businesses incorporate AI-powered automation solutions to address their labour-intensive work, many still fall short in meeting their goals as the solutions they have invested in are not suitable for processing unstructured data," he said.

"The adoption of integrated automation platform (IAP) solutions will future-proof business operations as unstructured data continues to grow – (we're) expecting it to reach about 80% of the world’s data by 2025. As we step into 2020, businesses should begin looking to adopt AI tools that are not just capable of automating certain functions within the business operations, but the entire process."

Jones said regional businesses are increasingly moving to implement 'digital workers', which he defines as specialised bots with skills that complement human workers’ responsibilities by taking on repetitive and time-consuming tasks. “Growth in integration with third-party cloud AI services by enterprises will encourage digital worker implementations, which can then help with creating end-to-end automation workflows through pairings to chatbots and other triggering mechanisms,” he predicted.

In the energy vertical, AI will be of great help as well. Mathias Steck, Executive VP, DNV GL – Energy, said, “The high amount of dynamic generation (wind and solar), combined with increasingly customised loads and multidirectional power flows (e.g. prosumers) adds complexity, beyond human ability to control.

“IoT and digital technologies will provide the basis (data) for AI to optimise the energy ecosystem (generation, transmission, distribution, loads) - a prerequisite for the energy transition.” IoT stands for the Internet of Things.

Pain before gain

Yang of ABC Technology said that some effort will be required to implement AI, however. “Business users need to go through some unavoidable initial investment of time to bring in the AI, (and must have) openness and flexibility to adopt new technology. Once they do that, they will boost productivity and their value to their companies,” he said.

Tan of JOS Singapore stressed the need for a proper AI implementation strategy. “A customer-centric strategy should be adopted throughout all technologies. Think about how you can streamline processes for frontline employees to better serve customers. For example, a tool like RPA can help to ease the workload for employees. And we all know happy employees make for happy customers,” he said.

“Many businesses attempt to manage risk when venturing into AI projects by creating proofs-of-concept. However, this thinking is inherently flawed because the selected use case is typically too small to ever justify the ROI. Companies need to think bigger and plan for the long term, while allowing space for small failures along the way,” recommended Sithasanan of Hyperlab.

“Businesses need to approach automation and AI projects as more than a data or IT project and design a blueprint for a holistic user experience. This means getting internal teams from marketing, branding, IT and customer experience to work together with partners that have the right skills and tools.”

He added that the design, data and business strategy for each AI project need to be clearly thought through in order to ensure that the results are viable and scalable over the long term. “Scalability is key and needs to be baked into the solution from day one,” he said.

Chuah Seng Heng, VP and GM, Asia Pacific and Japan, Motorola Solutions, had an ethical angle. “With the increasing use of AI in a number of industries, the right controls need to be in place to ensure AI is used responsibly, ethically and in a manner that supports human decision-making. Any tool used incorrectly has the power to cause harm, but applied in the right way, AI can bring considerable benefits to critical industries including public safety and law enforcement,” he said.

That thorny infrastructure problem

Source: Cloudera. Micallef.

Mark Micallef, VP of Asia Pacific and Japan, Cloudera predicts that AI factories will become popular in the Asia Pacific region. “Getting machine learning into production or embedded in the entire organisation can be challenging. Since businesses may be running on an inflexible, siloed IT infrastructure, it will be difficult for machine learning systems to access the necessary data, and scale as needed,” he pointed out.

“We believe more organisations in the region will build an AI factory in future to overcome that issue. Built on a platform that unifies and powers all machine learning workflows across multiple clouds and on-premise environments, an AI factory enables the process of building, scaling, and deploying enterprise machine learning applications to be automated, repeatable and predictable. Since this eliminates the complexity of deploying machine learning apps, organisations can easily operationalise and scale machine learning capabilities across the enterprise, which ultimately empowers the entire company to make data-driven decisions.”

Could AI make our lives worse?

That said, AI still seems powerful enough to replace us, especially if the creative aspect is within grasp as Appier has predicted. Stu Garrow, SVP & GM APAC at Talend says that AI must balance on a fine line between human work and automation.

“With the rise of automation to carry out day-to-day business functions, leveraging AI to augment human capabilities will continue to be a delicate balance,” he said.

Steck lists the risks: “Although AI will improve reliability, safety, and cybersecurity, optimise the system and enhance the customer experience, it also poses risks of data exploitation, prediction, and profiling. The inability to retrace how AI made the decisions is an added challenge.”

Source: LogRhythm. Wong.

Biased AI decisions could be a problem. Joanne Wong, Senior Regional Director for Asia Pacific & Japan at LogRhythm, said that in 2020, LogRhythm expects that an insider will manipulate AI to wrongly put an innocent person in prison.

“As humans, specifically data scientists and engineers, continue to train AI, AI has inadvertently adopted the same human biases we thought it would ignore. However, this hasn’t stopped the legal system from employing it,” she said.

“With AI already primed to make biased decisions based on the information it receives, an insider could exploit this to feed it false information to more directly implicate someone of a crime. In making AI more human, the likelihood that it makes mistakes will increase.”

Getting serious about AI

Source: Juniper Networks.
Bament.

Who is really offering AI? Sally Bament, VP of Service Provider Marketing, Juniper Networks says that everybody claims to have it, but that this is not necessarily true. “Everyone wants a hand in AI, and many emerging businesses have been guilty of 'AI-washing' rather than delivering true self-learning and operating smart technology,” she said.

“In 2020, the true AI providers will distinguish themselves from the imposters through capital investments and purchases. As more large telecom companies and enterprises look to utilise advanced AI capabilities for streamlined network operations and connectivity, we will see a spike in mergers and acquisition (M&A) deals targeting smaller AI startups in 2020. After all, AI and automation are critical to managing ever increasing network complexity and ensuring fast delivery of services in the 5G era.”

Explore:

RPA is also discussed in R is for robotic process automation, a part of the A-Z of technology predictions 2020, to be published in January 2020.