Why companies must use AI to think differently, and not simply to cut costs

Organizations have to look at how artificial intelligence (AI) can enable them to do things differently, rather than at a lower cost, in order to stay relevant in the future. 

In fact, 21st-century companies will not be defined by the quality or the price of their products and services, but by their use of AI, said Mike Walsh, futurist and CEO of tech consultancy Tomorrow. There will be significant shifts in the way these businesses operate in the future, said Walsh, who was speaking at ST Engineering's InnoTech Conference held this week in Singapore.

Also: AI is coming to a business near you. But let's sort these problems first

Future-oriented companies will move from building products and services to developing data-powered platforms that can be reapplied to adjacent markets, he said. Tesla, for example, has expanded beyond car manufacturing to provide insurance services, using real-time driving behavior to differentiate its offerings. 

Growing interest in generative AI also has pushed companies to tap their data assets and build their own large language models. This, he added, will create new proprietary data platforms that will power new types of products and services. 

Rather than focus on how AI can help them do what they already are doing faster and cheaper, organizations need to think about how the technology can do things differently, Walsh said.

AI will lead to changes across four key areas encompassing scale, speed, sustainability, and scope, he said. Businesses will have to figure out the right "talent density" or workforce they need to drive their operations and be able to respond quickly to market opportunities. 

However, powering the large language models that AI vendors -- including Google, Meta, and OpenAI -- are building will require massive computing capacity. Such increased demand underscores the need to ensure data centers are sustainable and fully powered by renewable energy sources, he noted. 

Also: Generative AI will soon go mainstream, say 9 out of 10 IT leaders

Foremost, organizations must widen their scope. If they are not looking beyond using AI simply to do what they already are doing more cheaply, they are missing out on what is possible, Walsh said. 

And when leveraged well, AI can fuel both businesses and economies. It will cause significant disruptions, though, which have to be managed. 

AI is projected to increase global GDP by 7%, or an estimated $7 trillion, over 10 years. At the same time, it can expose two-thirds of job roles to some degree of AI automation, said Janil Puthucheary, Singapore's senior minister of state for communications and information, citing the forecasts from a Goldman Sachs report.

Cybersecurity attacks also have climbed amid accelerated digitalization, with Singapore clocking 132 reported ransomware cases last year, Puthucheary said during his speech at the conference.

Also: Generative AI and the fourth why: Building trust with your customer

He noted that local organizations need to be equipped to manage these disruptions and turn them into opportunities and competitive advantages. 

Singapore, for instance, is focusing its efforts on strengthening the foundation to support its digital infrastructures, including hardware and software, he said. He pointed to initiatives such as its Digital Connectivity Blueprint, which includes plans to boost local connectivity to 10Gbps within the next five years and double submarine cable landings within a decade.  

These aim to ensure organizations can better leverage latency-sensitive technologies such as AI and cloud, he said. 

Policies outlined in the country's Cybersecurity Act 2018 also are under review to identify potential areas that need finetuning, so the right safeguards are in place for these digital infrastructures, he added. 

Also: Leadership alert: The dust will never settle and generative AI can help

Acknowledging that Singapore is one of the most sophisticated global smart cities, Walsh said the Asian market offers a good staging ground to create platforms to do things differently and that cannot be easily replicated elsewhere. 

"If you are using ChatGPT to write email for colleagues who, in turn, task a chatbot to read and summarize your messages, there is a strong possibility your organization may have missed the real point of the AI revolution," he wrote in a recent LinkedIn post. "Generative AI is neither a toy nor a fad. It is a provocative invitation to consider what the real value of human work is."

"The most forward-thinking organizations will ignore basic productivity gains from generic AI tools and, instead, explore training proprietary large language models with their own unique datasets," he added. "Only by making generative AI technologies your own can you unleash both the hidden power of your intellectual property and organizational knowledge, as well as empowering your people to do more interesting, creative, and differentiated work."

ST Engineering, for one, hopes to help its customers do so with the launch of the first of several generative AI tools. 

Also: AI has the potential to automate 40% of the average workday

Called AGIL Vision, the video analytics system can perform searches based on objects, colors, actions, or human behaviors, and distinctive clothing marks such as brand logos. For example, it can seek out someone in a shopping mall wearing a blue shirt that bears a specific logo or detect people who are smoking or involved in a brawl. 

The system is trained on various large language models and natural language processing, and can search for images without meta tags. Because it carries out searches based on objects or actions that fit the specified keywords, not by a person's identity, AGIL Vision addresses any concern companies may have about privacy, said an ST Engineering spokesperson. 

The video analytics system can be deployed over a private cloud or on-premises, the latter of which would ensure data is transmitted within a local corporate network. It encompasses a box that connects to the company's CCTV system.