Aaron Tan2024-01-10 17:23:39Computerweekly
There is no doubt that 2023 is set to be the year of generative artificial intelligence (GenAI), which has caught the world by storm through its wide-ranging impact on business and IT. In the new year, more organisations will try to operationalise the technology, manage its risks and scale up adoption.
While GenAI is not expected to supersede other IT priorities such as cloud, application development and cyber security, it will augment those areas by making it easier to build software and improve cyber resilience, among other outcomes. In this round-up, we review some of the key IT trends, including GenAI and other developments, that are likely to shape the industry across the region in 2024.
Following the explosive growth of GenAI this year, the new year will see the introduction of smaller and more specific large language models (LLMs), said Gavin Barfield, vice-president and chief technology officer for solutions at Salesforce ASEAN.
While foundational LLMs will be the backbone of GenAI, they are generic, designed for a mass audience, and expensive to run. With these constraints, more companies will start developing and using a combination of smaller, domain-specific language models for cost, performance, and latency reasons.
These specialised LLMs will be trained on specific datasets that can provide GenAI assistance in different and unique industries. For instance, a specialised model for the medical tech industry in Singapore will contain a wealth of data focused on medical knowledge and terminology, grounded in the local context.
“Across ASEAN, we may also see models fluent in local languages, such as those that can generate content in Bahasa Indonesia, Thai or Vietnamese. Specialised models will not need to store large amounts of data for general knowledge questions, and the smaller database will be more cost and energy-efficient to run for specialised industries,” Barfield said.
On software development, Barfield noted that AI will drastically change the way software is built. In the past, IT teams built software applications by connecting clicks to code. When the end-user clicks on a button to execute the code, the data that connects to the code will be augmented accordingly. Barfield said this approach results in applications that are pre-built with a fixed set of features, which can be limiting for customers.
Building with AI is completely different. Not only is the user interface much simpler, but teams can now build in natural language with just a set of instructions, otherwise known as a prompt. The prompt will then connect to the LLM and data that sits below it to generate the desired output.
“With data grounding, the way we build will be a lot simpler given that we are working with contextual data, which ensures that the output is accurate and relevant,” Barfield said.
And with the democratisation of AI, even teams in non-traditional IT roles can become digital builders. Using low-code or no-code tools and coding prompts, business users can customise their own workflows and systems, and pull together their own features to meet specific needs, he added.
Across ASEAN, we may also see models fluent in local languages, such as those that can generate content in Bahasa Indonesia, Thai or VietnameseGavin Barfield, Salesforce
But Kevin Reed, chief information security officer of Acronis, warned that with LLMs being used to generate code, hard-to-diagnose bugs and security vulnerabilities may arise and be exploited with or without LLM assistance.
“This includes ransomware authors who will use LLMs to develop malicious software. Because it’s hard to deduce the intent of the software development, no matter what protections LLMs will try to put in place, there will always be bypasses,” Reed said.
Recent incidents involving prominent organisations such as Singapore Airlines, Singtel, Singapore’s National Trade Union Congress, Microsoft, SolarWinds and Asus have highlighted the severity of software supply chain attacks.
In APAC, more than 50% of companies have been negatively affected by between two and five cyber security breaches in their supply chain, according to Armando Dacal, group vice-president at DigiCert Asia-Pacific and Japan.
In the new year, Dacal expects the software supply chain to become more robust, with inspections at various points of delivery. The composition of embedded software will become more transparent with the increasing adoption of software bills of materials (SBOMs).
The cloud infrastructure market is being dominated by global hyperscalers that offer a sheer number of cloud services ranging from cloud databases and compute services to application platforms. Yet, these hyperscalers do not typically offer services customised to specific geographic regions’ requirements and unique use cases, expecting local customers to adopt a “one-size-fits-all” approach to cloud computing.
Sunny Chua, Singapore general manager of Wasabi Technologies, expects specialised providers that offer more cost-effective storage solutions to gain ground, challenging the dominance of hyperscalers – especially as ASEAN anticipates generating massive volumes of data in its pursuit of a $1tn digital economy. The demand then arises for specialty cloud storage providers capable of delivering predictable, secure, and scalable storage, coupled with seamless and rapid access to data.
The growing interest by governments in protecting sovereign interests through various data sovereignty schemes can impact how organisations store, process, use, analyse and share data in different jurisdictions. Additionally, AI systems like GenAI, have raised issues of sovereignty during the training process.
Against this backdrop, 2024 will see the rise of the sovereign cloud, as more organisations respond to government regulations requiring them to store data within their own jurisdictions for greater data control, said Matthew Oostveen, vice-president and chief technology officer of Pure Storage in Asia-Pacific and Japan. Countries that are taking the lead in building sovereign cloud capabilities include Australia, New Zealand, Japan and Indonesia.
There is no question that data can propel an organisation’s digital growth ambitions, enabling enhanced intelligence and efficiencies. Not to mention, for any business looking to jump on the AI bandwagon meaningfully, it is their data that will ensure accurate outputs that are tailored to the business.
At the same time, as businesses create, share, and store more data than ever before, more blind spots are also surfacing. Specifically, unstructured data that makes up 80% of businesses’ data is posing to be a significant data security challenge. Often scattered across various platforms, devices and repositories, unstructured data adds an extra layer of complexity in being difficult to monitor and secure.
Sustainability will need to step back into focus, with a dedicated commitment – from individuals to entities – on reducing the watts we consume to save power for others in our datacentre regions and ensuring that we are scaling efficiently for the years to comeJonathon Dixon, Cloudflare
To mitigate the risks associated with unstructured data, organisations must evolve their identity security approaches to include managing access to this class of data, said Chern-Yue Boey, senior vice-president for Asia-Pacific at SailPoint, calling for the use of automated identity security solutions to address identity-related threats.
Sentiment analytics empowers businesses to track and to react to the emotional state of customers at each stage of their journey. While this is not new, the progress in technology such as AI is making it easier to detect such emotions in real time, said Angie Tay, executive vice-president and group chief operating officer at TDCX.
Tay said such insights can be used to provide tailored responses to individual customers, such as whether to channel them to automated channels or a human agent. Being aware of whether a customer is enthusiastic about a recent purchase or frustrated due to poor customer service allows for more informed choices, such as determining whether it’s the right time to ask them to leave a review or reach out to make amends.
“A deeper understanding of how sentiment affects loyalty, satisfaction and advocacy scores will be core to business growth strategies in 2024, helping to build brand preference and loyalty,” Tay said.
Ransomware attacks are relentless and will affect unprepared organisations. Most of these malicious acts involve a series of tactics and techniques known as the cyber kill chain. Already, threat actors are leveraging AI for attacks through the likes of FraudGPT and WormGPT.
Reuben Koh, director of security technology and strategy at Akamai in Asia-Pacific and Japan, noted that 2024 will see more cyber criminals automating the arduous tasks of target prioritisation, enhancing ransomware encryption to improve resistance against decryption and reverse engineering, and using AI chatbots to improve the scale and efficiency of ransomware attacks.
“For organisations, the call to action is clear: Fortify cyber resilience by getting holistic visibility and embrace zero-trust access and segmentation through meticulously verifying all application access,” Koh said.
Hyundai’s small-scale electric vehicle manufacturing plant in Singapore may be the ultimate smart factory, but the reality is that many businesses can’t start with a greenfield site when building a new facility. While 2024 will see significant growth in smart factories, this will come from existing factories becoming smarter, according to Mei Dent, chief product and technology officer at TeamViewer.
Dent said having operational data on the factory floor will become a crucial step for organisations to leverage data analytics to drive efficiencies and build industry specific solutions for smart factories. “The technologies to deliver smart factories exist and can be integrated with existing operations, meaning that a new build is not always necessary,” she added.
In addition, augmented reality (AR) will be used more on the factory floor, Dent said. “Through smart glasses and AR software, the person on site can be walked through processes by an expert based anywhere. This will deliver increased uptime to the factory and enable employees to optimise their processes”.
The AI hype cycle significantly slowed sustainability efforts in 2023. Skyrocketing demand for compute and AI, global conflict in regions that supply oil and gas, and unexpected extreme weather events will continue to drive up the cost of electricity, said Jonathon Dixon, vice-president and managing director of Cloudflare in Asia-Pacific, Japan and China.
“Sustainability will need to step back into focus, with a dedicated commitment – from individuals to entities – on reducing the watts we consume to save power for others in our datacentre regions and ensuring that we are scaling efficiently for the years to come,” he added.
Terry Maiolo, OVHcloud’s Asia-Pacific vice-president and general manager, underscored the pressing need for greener datacentres to address the growing carbon footprint associated with the use of emerging technologies like AI.
“Power, water and carbon usage are all critical metrics in assessing the efficiency of datacentres. Fortunately, water cooling technology has proven itself as a far more energy-efficient and sustainable way of cooling, eliminating the need for air conditioning – with significant benefits to costs and environmental impact.
“Collaborating with efficient providers, especially in light of the increase in AI workloads, becomes increasingly valuable. Such partnerships can pave the way for a clear, bright trajectory toward a more environmentally responsible future,” Maiolo said.
With more pressure on Australian directors to take more responsibility for their organisation’s cyber resilience, 2024 will likely see a shift in corporate boards where cyber security knowledge becomes a meaningful part of the balance of a board’s expertise, said Michael Armer, chief information security officer at RingCentral.
“Over the next year, corporate boards will prioritise onboarding members that combine cyber security experience with business acumen,” Armer said. “From securing valuable IP [intellectual property] to privacy and trust issues to reputation management, cyber security has evolved into a board-level issue.
“While not every board will look to include a dedicated cyber security expert, we will see boards collectively turning to individuals with the requisite security and risk expertise to provide much-needed oversight,” he added.
Declare：The sources of contents are from Internet，Please『 Contact Us 』 immediately if any infringement caused
