The rapid advancements of computing power have paved the way for AI’s integration into everyday devices and services 

by AZALEA AZUAR & RADZI RAZAK / Pic b BLOOMBERG

IN THE not-so-distant past, artificial intelligence (AI) was considered a fascinating concept confined to the realm of science fiction. However, over the past decade, this once elusive technology has managed to break free from its theoretical confines and firmly establish itself in the mainstream. 

The journey of AI into the mainstream can be attributed to a multitude of factors and chief among them is the rapid advancements of computing power. The exponential growth of processing capabilities has enabled AI algorithms to process vast amounts of data in real time, opening up a myriad of possibilities. 

This newfound power has paved the way for AI’s integration into everyday devices and services, making it an indispensable part of our modern lives. 

Another pivotal factor in AI’s mainstream success lies in the increasing availability of big data. The proliferation of Internet-connected devices and the rise of social media have generated an unprecedented volume of data. AI algorithms thrive on data, using it to learn, adapt and make intelligent decisions. With the vast amounts of data readily available, AI has found ample fuel to refine its abilities and deliver valuable insights across a range of applications. 

As a result of these converging factors, AI has managed to break down the barriers that once separated it from the mainstream. From voice assistants on our smartphones to personalised recommendations on streaming platforms, AI has seamlessly integrated into our daily routines. 

However, it is important to approach the integration of AI in the workforce with a balanced perspective. 

According to the Malaysian Research Accelerator for Technology and Innovation (MRANTI) CEO Dzuleira Abu Bakar, AI can enhance human capabilities, create new jobs and increase productivity if approached with balance. 

“Industry pundits have said that the future of work is likely to involve a combination of human and machine collaboration, with AI serving as a tool to support and enhance human work rather than entirely replacing it,” she told The Malaysian Reserve (TMR). 

Dzuleira noted that Malaysian organisations face challenges in AI adoption due to limited expertise and funding. However, leveraging advanced technologies can address societal and sectoral issues by enabling technology-driven solutions. 

She highlighted the benefits of AI in the workplace, including task automation and the transformation of job roles to augment human capabilities as AI can enhance efficiency and productivity in medical diagnosis, legal analysis and customer experiences. 

Furthermore, she added that AI has the potential to generate new jobs and industries, although it still relies on human skills. 

“Historically, technological advancements have led to the creation of new industries and jobs. 

“As certain tasks become automated, new job opportunities emerge in developing, managing and maintaining AI systems,” she said. 

Although AI excels in data tasks, it lacks human skills such as emotional intelligence and creativity. 

However, she advised that there is a need to train the youth to manage AI as soon as possible. 

This is by highlighting individual strengths and interests instead of standardised knowledge. 

“Displaced workers may face challenges in transitioning to new roles, requiring upskilling or retraining programmes to ensure their employability. 

“Policymakers, businesses and societies need to address the potential impact on income distribution, labour markets and the overall welfare of individuals,” Dzuleira added. 

AI Won’t Harm the Labour Market

As for AI5 Technology founder and CEO Boris Yeong, he believed AI will not harm the labour market because it is unable to compete with the actual “human brain”. 

“At this moment, we will talk about the next five years. I don’t think AI machines can be equivalent to humans. 

“So, if you or other people say it is harming the labour market, you don’t have enough value to start a company using AI,” he told TMR. 

Yeong’s company provides AI solutions to insurance companies such as image recognition services. 

“You need to teach the machine how to recognise this. 

“For example, in car insurance, we teach the machine how to recognise the car bumper, whether it is dented,” he explained. 

Malaysians Happier with AI 

ChatGPT’s launch in November last year sparked a debate on generative AI’s impact on jobs. 

This issue is now a concern in Malaysia, as the pandemic has led larger companies to accelerate the deployment of chatbots and other AI tools. 

The National Tech Association of Malaysia (Pikom) Research committee said generative AI is seen globally as a threat to knowledge jobs, similar to how automation affected manual labour. 

However, it can also bring in more jobs. 

“Malaysian and Malaysian-based companies, particularly the larger and more prominent ones, have been progressively incorporating AI into their processes and procedures. 

“They include telecommunication entities such as Maxis and Celcom, financial institutions including the like of Maybank and CIMB, e-hailing and e-delivery platforms as Grab and e-commerce marketplaces Lazada and Shopee,” it told TMR. 

On the other hand, smaller companies use AI due to pandemic restrictions and cost-cutting. 

The committee also expects AI to spread across Malaysian businesses in five years. 

Microsoft’s Work Trend Index 2023 shows 62% of Malaysian employees are worried about being replaced, but over 80% are willing to use AI to lessen their workload. Managers in Malaysia are more favourable towards AI for productivity gains without reducing staff.

“Pikom understands that while some companies recognise the value of AI in terms of cost savings and efficiency, there are still many that are either ignorant of the technology or remain sceptical over its implementation. 

“Yet, Pikom predicts most firms will adopt AI once its benefits prove evident,” it noted. 

List of Everyday AI Apps 

Currently AI has been advertised with more apps instead of hiding behind services. This would help an individual themselves to control and dictate how AI could be used in their daily lives. 

Whether one is a freelancer or professional, these AI platforms are available for the public and would help enhance their productivity at work. 

Some of these apps include: 

Otter.ai 

Otter uses AI to transcribe audio into text, ready to read, highlight and copy. 

Users can install the app (which is available for both Android and iOS) to record on the go or convert audio and video files to text. 

They can also set up Otter Bot for automatic transcription during video meetings, but should be aware of potential privacy concerns. 

Developers have included strong privacy features such as SOC 2 Type 2, GDPR, CCPA and VPAT compliance. 

The AI platform can also be used as a collaboration tool which summarises conversations for teams, making recall and collaboration easier. 

Otter integrates with Zoom, Microsoft Teams and Google Meet for meeting note capture. 

Moreover, the platform is launching a new service called OtterPilot which is designed to automate meetings and help professionals save time and increase meeting productivity. OtterPilot sends AI- generated meeting summaries with hyperlinks to invitees or Otter groups via email. 

ChatGPT 

ChatGPT is an AI-driven tool that mimics human conversation and helps users with tasks like email, essay and code writing. 

The software was launched on Nov 30 last year, and is the mastermind behind art generator DALL-E 2 and automatic speech recognition system Whisper. 

ChatGPT employs specialised algorithms within its Generative Pre-trained Transformer (GPT) to detect patterns in data sequences and utilises the GPT-3 language model, a third-generation neural network machine learning model. 

It also employs transformer neural networks with deep learning to generate natural language from vast amounts of data. 

The transformer predicts text based on training data sequence. 

It is currently free for public use during its research phase but a paid version ChatGPT Plus was launched in February. 

OpenCV 

Open Source Computer Vision (OpenCV) is a vast open-source library for computer vision, machine learning and image processing. 

It is crucial for real-time operations and can identify objects, faces and handwriting. Using the programming language Python, it has libraries like NumPy which can process OpenCV arrays for image analysis by using vector space and mathematical operations to identify patterns and features. 

In object identification, the model distinguishes objects by identifying specific instances. 

The programme was launched by Intel Corp in 1999 and was originally written in C/C++ but is now commonly used in Python for computer vision. 

The first alpha version was released in 2000 at the IEEE Conference on Computer Vision and Pattern Recognition, followed by five betas between 2001 and 2005. 

DALL-E 2 

Dall-E 2 is an AI that generates images from text prompts where it creates new images in various styles based on user input. 

The name Dall-E 2 is a nod to merging art and AI, with DALL hinting at Salvador Dali and E at Disney’s Wall-E. 

The name combo represents the abstract and surreal illustration produced by machine automation. 

Dall-E 2 was launched in January 2021 and uses deep learning models and GPT-3 to generate images from natural language user prompts. 

It is an evolution of OpenAI’s Image GPT from June 2020 and uses a neural network to create high-quality images and extends the concept to generate images from text prompts like GPT-3 does with text. 

Blender 

Blender is a comprehensive open-source 3D creation tool. 

It has powerful capabilities for 3D creation, including modelling, texturing, rigging, animation, lighting and more. 

This AI tool is great for static models or animation as it also includes a video editor and game engine for more capabilities. 

Blender was first created by Ton Roosendaal in 1994 for his animation studio. Originally, it is not an AI tool but as of this year, it has incorporated AI through “Stability for Blender” so users can simply generate 3D models using the software’s AI, allowing users to create AI textures, effects and animations without artistic skills.

However, Stability for Blender needs  API key and Internet, but it’s free and no software or GPU needed. 

---

• This article first appeared in The Malaysian Reserve weekly print edition