Feature Articles

Should you be worried about Generative AI? Local tech dignitaries say its time to embrace it.

By Vijay Anand - 18 Jun 2023

Should you be worried about Generative AI? Local tech dignitaries say its time to embrace it.

Note: This feature was first published on 1 June 2023.

(Image source: Illustrated with Microsoft Bing Image Creator, powered by DALL-E generative AI engine)

Ready or not, AI is here to assist you

You can't deny that the recent buzz around AI and Generative AI (GAI) has grown exponentially. Be it companies launching new services (such as Shutterstock's AI image generation platform) using prompts (just like ChatGPT), or working on similar services but tailored with in-house domain-specific data (such as utilising NVIDIA AI Foundations) or about the ethics and risk in embracing AI (you can hear more from the CEO of OpenAI, maker of ChatGPT); there's no shortage of the ups and downs of our newfangled obsession.

To get a better perspective of why AI is in such a big spotlight these days, it's essential to understand sentiments from workers, and just earlier this month, Microsoft released their comprehensive Work Trend Index (WTI) annual report after surveying 31,000 full-time or self-employed workers across 31 countries.

Focus Time is becoming a rare commodity

While 'time' as a commodity is debatable, this is best put into perspective by Jacob Carlile, that time is the commodity we trade every day in exchange for value. Here's a further snippet:-

It's not about how much time you have; it's about what you do with the time that you DO have that defines your actions.

One of the most critical findings from Microsoft's WTI report is that 2 in 3 respondents mentioned they grapple with time to do their job. And the key takeaway is that 68% said they don't have enough uninterrupted focus time during the workday. In Singapore, this figure shoots up to 82%. We hope it's not because of our EDMW sub-forum (pun intended)

Apart from increased social media distraction, when you narrow it down to the cause of the issue, it's easy to understand where all of our energy is being siphoned these days -- communication. What used to be more structured meetings to carve one's time clearly, the pandemic has skyrocketed the use of communication tools such as group chats and virtual meetings to get across messages instantly, clarifications and ideas sooner. This robs us of our work and creativity time with more interruptions, not to mention follow-up emails and the next steps to be taken. Even Microsoft's own Work Trend Index report cites 57% of their surveyed users spend time on communications (catch-ups, emails, Teams Meeting, Teams Chat, etc.). In comparison, only 43% of their time goes towards creating content (or work) using Office applications like OneNote, Excel, Word, etc.

Time spent in a workday by a typical Microsoft Office user. (Source: Microsoft Work Trend Index 2023.)

Microsoft terms this problem as the Digital Debt that's costing innovation because people are disrupted and have less focus time. In fact, in the same survey, 2 in 3 leaders attributed that this has reduced noteworthy ideas within their teams as a concern.

And the solution, it seems, is AI.

What can AI or Generative AI do to help?

While there is still the underlying concern that AI could replace their jobs, WTI's findings reveal that a vast majority would prefer to offload as many tasks as possible to an AI and reduce their workload. Globally, 70% responded positively to AI uplifting their jobs, but this figure goes up to 80% for employees in Singapore.

Simply put, AI holds the key to increasing employee productivity by improving and automating mundane but repetitive tasks such as condensing reports or even writing one up based on available data points to be interpreted with domain-specific parameters and knowledge applied to it. AI has now become an essential tool to unlock new workflows and achieve more for one's organisation, as captured by WTI's survey, where one of the questions asked was which of the following would bring the most value to one's workplace:-

In the WTI survey, one of the questions asked was which of the following would bring the most value to one's workplace, and an overwhelming number of responses were efficiently related to enhancements to boost productivity. The graph sample is global, or 31,000 workers.

Are we ready for the Generative AI boom?

Earlier this month at Marina Bay Sands, Salesforce put on their World Tour Essentials event. It held an interesting panel discussion with local dignitaries such as Terence Chia, Cluster Director for IMDA and Assoc Prof Damien Joseph, Associate Dean for Nanyang Business School at NTU, on the rise of AI and its impact on the workforce.

According to Salesforce's latest Digital Skills survey conducted in February 2023 across 11 countries with over 11,000 respondents, only 1 in 10 workers say they possess AI skills. In Singapore, a thousand people responded to this survey, and 15% said they currently possess AI skills.

When asked how this statistic will change the workforce and skills needed, Terence Chia agreed that while there is a shortage of tech talent, there's not much AI evolution taking place on the local stage. Contrary to the findings, Chia mentioned that AI is actually used almost everywhere today in our everyday routines. Fire up Netflix, and you will notice the video streaming engine learns from your past viewing preferences to offer more of the genres you've been watching. Likewise, your social media feeds are also customised to your viewing preferences and help eradicate surfacing fake news. Google's predictive search algorithm (autocomplete), autocorrect functionality, digital assistants, smart home devices, recommended buying lists from your preferred shopping platform (such as Amazon, Lazada, Shopee, etc.), and the list goes on where AI systems and recommendation models are silently working in the background, improving our everyday experiences. 

You don't need to know how AI works; learn how to use it.

According to Chia, the more pertinent question is if you know how to use AI effectively. People don't need to know how the tech behind it works or how to code it, but it is increasingly getting essential to know how to harness it to improve our work and lives.

Asking the right questions, or prompt engineering, would be the path to using AI effectively and deriving the intended outcomes. This is no different from knowing how to work the search bar on Google or any other search engine to narrow the resultant output to something genuinely beneficial. In the same vein, we'll have to gain new skills to learn how to use AI or GAI for our gain. Sectorial use cases will vary from generic uses; hence it will be a while before specific industries and niche verticals learn how best to embrace them to uplift their work.

Growing one's AI aptitude

Professor Damien also shared that his students have started experimenting and using ChatGPT during their work. Experimenting and learning the boundaries of this AI tool are important elements moving forward, and he feels people shouldn't be afraid of it. Knowledge is, after all, power, and not knowing what the system is capable of doing is far worse than knowing the boundaries to make the best use of it. People have become accustomed to the search engine, and it's time to learn the art of asking the right questions to an AI prompt. He reiterated that people shouldn't be hung up on skills required in the past, such as how to code the system (while still relevant for those keen on engineering the system), for example, but instead move on to what matters more, such as understanding how to generate preferred outcomes that will assist many organisations and improve workflows.

In some other implementations of collaborative AI, we may lose some decision-making abilities. Workers should not fear the latter because the antidote to skills loss is upskilling to higher order abilities and toward expertise. -- Assoc. Prof. Damien Joseph, Associate Dean (Undergraduate Education), Nanyang Business School, Nanyang Technological University.

About academic integrity concerns with GAI

As shared by Profession Damien, discerning what's written by GAI and what's produced by oneself is the most significant ongoing concern for adopting GAI openly in academia. Yet, it can't be disregarded. It has to be acknowledged that an AI assistant exists at everyone's disposal, and students need to be educated on its usage. He shared that contextualisation of information in any written work is key, and for now, GAI, in general, is poor at proper referencing and accuracy. As ABC News describes, it's a powerful study buddy, as one college in Australia has shown how it can be used effectively.

How can companies adapt?

When asked what companies must do for this transformation, Chia cited how the world pivoted to embrace remote and hybrid working due to the Covid outbreak. Similarly, companies need to figure out what aspects and processes where AI can help their business the most and work towards defining this.

While large companies can get partnerships and develop AI in-house, SMEs face the challenge of cost to make these implementations. Chia shared they would need an equivalent of a CTO-on-tap or as a service  (CTOaaS) to bridge the gap.

(Image source: Unsplash)

Beyond charting the path and working out the cost benefits, Chia stressed that the next most important question is if the company has the right people. Small companies can respond better and are inherently more agile as culture change becomes easier. However, this is difficult for large companies, and a shift in focus and communicating this both internally and outside the company is a bigger necessity.

All in all, it's crucial to have the right people, tech and culture in place to embrace the change.

Regulation and data protection concerns through the use of AI?

Data privacy and how personal data is managed have been an ongoing concern, and the use of AI naturally brings up a more significant concern. Chia assured that a data protection officer (DPO) is a must in all organisations to comply with PDPA requirements to ensure customer data is managed and used in the proper manner. He encouraged organisation heads to continue having conversations with the industry folks to collectively evolve the needs of personal data protection in the right way and improve standards where necessary.

(Image source: Unsplash)

The need for guardrails in the ethical use of AI is also an important aspect of development to trust an AI system, such as how NVIDIA has come up with its NeMo Guardrails open-source software to help developers guide generative AI applications to create responses that are accurate, appropriate, on-topic, and secure.

Similarly, establishing a national framework will go a long way to help businesses operating in the country assure their residents quickly that an organisation's AI service meets a minimum level of service and transparency guidelines. To that extent, IMDA's A.I. Verify was launched last year to help establish an AI governance testing framework and toolkit that is currently in the testing and collaboration (MVP) phase. The overarching idea is to help assure people that an AI service employed follows a proper set of operational principles and rules and check that it works as per its marketed function.

Talent Acquisition and Training in the Age of AI

(Image source: Unsplash)

Insights from Salesforce's latest Digital Skills survey indicate that 97% of global workers believe businesses should prioritise AI skills in their employee development strategy. As such, not only are people ready to utilise AI to their benefit, but they are also open to gaining the required skills. In that same survey outcome, Salesforce shared that the majority of leaders in Singapore (97%) believe that the practice of skills-based hiring provides business benefits such as improved talent retention.

In the Microsoft WTI survey, managers and leaders were asked which of the following skills would be most essential for employees to learn to evolve with changes in tech advancements due to AI:-

  • Flexibility: To rapidly adjust to AI's integration in the workflow
  • Emotional Intelligence: To determine when to leverage a human capability instead of an AI capability
  • Analytical Judgement: To determine when to leverage an AI capability instead of a human capability
  • Creative Evaluation: To evaluate content produced by AI
  • Intellectual Curiosity: To ask AI the right questions
  • Bias Detection and Handling: To evaluate AI fairness in decision making
  • AI Delegation (prompts): To direct AI with the right prompts

This is the outcome of the question as captured in the infographic below:-

Clearly, everyone needs a lot of skills in their toolbox to approach AI with the right mindset, and there's room to grow even if AI is also eliminating certain job functions and repetitive work. Learning and development are critical drivers to moving forward.

Tech employers need to focus more on skills-based hiring, and less on traditional qualifications. This mindset shift will be key in ensuring our workers and companies remain competitive, and able to reap AI’s benefits. -- Terence Chia, Cluster Director, Digital Industry and Talent Group, Infocomm & Media Development Authority (IMDA).

On this note, Chia cautioned that it is difficult to plug people in and out of specific job functions, and it is thus vital to understand what aspects of AI can help guard, managed and evolve existing institutional knowledge. Profession Damien also underscored the importance of upskilling to evolve with times as tacit, institutional knowledge was once challenging to capture, but GAI has the ability to learn this over time through training and improved inferencing.

Recent new tools from Microsoft and Salesforce to enhance workflows

To further treat AI as the workplace co-pilot, Microsoft recently introduced their Microsoft 365 Copilot solution as part of its early access program, which will eventually be rolled out to its subscribers. It consists of the following:-

  • Copilot in Outlook to provide writing suggestions and communicate effectively.
  • Copilot in Viva Learning to create personalised learning journeys complete with resource discovery and scheduling time to upskill.
  • Copilot in Whiteboard to help carry out more effective Teams meetings by organizing ideas and summarizing meeting outcomes.
  • Copilot in OneNote to help users generate ideas, create lists and organize information.
  • Copilot in PowerPoint to create custom images (through DALL-E image creator) to support presentation content.

Salesforce too introduced AI innovations to help the business deliver more personalised customer experiences at scale in the form of their new Einstein GPT and Data Cloud capabilities are added to Flow, Salesforce’s portfolio of automation tools. By powering Flow with Data Cloud, this unifies customer data across channels and interactions into single, real-time customer profiles.

Meanwhile, Slack GPT is a new conversational AI experience natively integrated into Slack to help everyone in the organisation become more productive. Lastly, Tableau GPT and Tableau Pulse, taps into generative AI, to help everyone better understand and interact with their data. This allows users to surface new insights in a conversational way by simply asking questions within the console and help power smarter data experiences with contextual reports.

To help workers build their digital toolbox with new in-demand digital and soft skills, as well as learning Salesforce, the company is committed to building such capabilities and enabling users through their online academy and Trailblazer commmunity.

Excitement for AI and generative AI solutions is at an all-time high within the Singapore workforce. As more companies adapt AI and automation to boost efficiency in their operations, the need to train their employees to leverage these tools will become even greater. -- Sujith Abraham, Senior Vice President and General Manager for Salesforce ASEAN.

The Answer is Clear

Just as how the early stage of robotics and AI transformed the factories of today and required a vastly different skillset from traditional laborious roles, generative AI will also transform jobs and roles of today to make us more effective. Job displacement isn't totally unavoidable, but if the past is a good gauge, new doors and opportunities are rife, and it's up to us to embrace change and put GAI to good use or oversee the accuracy of the tools related to the trade. As the old saying goes, change is the only constant.

Join HWZ's Telegram channel here and catch all the latest tech news!
Our articles may contain affiliate links. If you buy through these links, we may earn a small commission.