For jobs to be created, new industries need to be created and for that, you need people. A unique opportunity for job creation is emerging with the rise of artificial intelligence (AI), which is set to not only become a new growth area of its own but will also upend innumerable existing industries.
AI is also set to drive a new wave of industrialisation and represents a seismic shift on the scale of the Industrial Revolution, electricity, and the internet. But how countries will approach this revolution will depend on the unique challenges each is trying to solve, and the ways in which AI can assist. In South Africa, for example, the problems government and the private sector choose to prioritise and apply AI to will be those that lead to new industries.
While some industry experts link the increasing adoption of AI to the emergence of tools and services aimed at simplifying AI implementation by addressing data challenges, enhancing integration, and ensuring data privacy, there are more significant driving factors at play.
The manufacturing sector, in particular, is finding AI to be a powerful tool for myriad applications and for a wide range of phases of the production process. Machine learning (ML), a branch of AI, is the most widely used application in the manufacturing sector, enabling companies to modulate the production process and enhance quality. Research studies by Capgemini show that there is an increasing trend of AI uses in the manufacturing sector globally, where nearly 29% of use cases are observed in maintenance and 27% in quality control.
Overall, global surveys indicate that 60% of manufacturing companies have embarked on using AI in their businesses to enhance product quality and achieve faster production in addition to the significant benefits the technology offers, regardless of the type and nature of business operation. The COVID-19 pandemic created further incentives for manufacturers to employ AI more intensively, prompting a shift toward AI-enabled operations across many companies.
AI can help predict maintenance schedules, improve productivity and quality, create custom designs, adapt market strategies, and improve supply chains. It can also reduce human error, assist with decision-making, free staff from repetitive tasks, help with troubleshooting, and improve safety.
How it will affect different industries
Language is key
Language is embedded in every process of industrialisation, which is why revolution in how language and AI intertwine is so significant. Large language models (LLMs) are powering tools like Open AI’s ChatGPT, which are already changing how people work by helping them summarise text, rewrite code, and dispense with repetitive tasks.
Natural language processing (NLP) of the sort used in smart speakers and chatbots enables AI to parse, interpret, and act on human voice commands. This means instead of humans having to create languages for AI, it’s able to adapt to them. With NLP systems being trained to recognise various languages, it also allows AI to go beyond the confinement of a single language, and to be customised to different abilities and problems, and different industries.
LLMs and NLP are already being used to create content, to power chatbots or train them to be better, to serve as an alternative to search engines like Google, for translating text, and to augment developers code-writing skills. They’re also helping with training. One study found that AI enabled customer contact centre workers to gain the equivalent of six months of experience in only two months.
Widespread adoption of technology hinges on its capacity to be tailored for specific projects, accommodate multiple languages beyond English, and comprehend a user’s query or command intentions. For instance, in the case of NLP, it can utilize advanced “intent classification” to automatically discern the purpose behind a question or comment, enabling chatbots to provide users with precise and prompt responses.
One of industrialisation’s greatest achievements has been its ability to automate labour-intensive or repetitive tasks. AI has the power to do the same in today’s data-driven corporate landscape, freeing data scientists from rote tasks and enabling them to focus instead on more accretive activities, like analytics, model building, and experimentation. But for this to work, businesses must explore AI solutions that automate the tedious data collection and sorting tasks crucial for enabling AI initiatives.
Trust played a pivotal role in the success of innovations during the Industrial Revolution, facilitating commerce and enabling the widespread adoption of automated manufacturing and international trade. Trust in product quality reduced the need for face-to-face interactions between customers and producers. In the AI revolution, trust focuses on two key aspects: the responsible handling of personal data and the reliability of AI algorithms.
IBM’s Global AI Survey found that nearly 80 percent of over 4,500 respondents identified the assurance of “fair, safe, and reliable” AI outputs as a critical factor in their technology adoption. This underscores the contemporary significance of trust in maintaining strong connections between AI technology and its users. As trust grows over time, the relationship between people and AI improves, meaning it can be used for a growing number of purposes.
Technical considerations for AI in industry
The integration of AI has brought about significant transformations in the manufacturing industry, offering numerous benefits. AI-powered solutions enhance efficiency, production capacity, and competitiveness in manufacturing. This revolution encompasses various facets, from collaborative robots (or “cobots”) working alongside human employees to machine learning algorithms predicting demand trends.
Robotic Process Automation (RPA), meanwhile, automates repetitive tasks, which can save time and money and free up workers for higher-value and more satisfying tasks. Digital twins — virtual replicas of real-world tools or environments — can offer real-time insights and optimise up-time while enabling data-based decision-making.
AI can also enable so-called “lights-out factories” that run 24 hours a day with little to no human intervention, and which can be paired with ML algorithms that predict demand or adjust to fluctuations in energy costs or other variables to maximise productivity while minimising costs.
The Industrial Internet of Things (IoT) enables machinery, sensors, and devices, to communicate with one another, enhancing operational effectiveness, security, and productivity. Not every manufacturer will need or be able to implement all of the above, but even by harnessing a few of these tools, they can reshape their operations and their industries.
Further use cases
AI implementation will shape how new jobs will be created, and the AI value chain will create numerous jobs in the process as companies make AI products and execute AI strategies for various use cases.
McKinsey groups AI use cases into 15 domains: Personalization, Sales Channel Optimisation, Digital Marketing, Integrated Supply Chain Optimisation, Robotics and Workforce Automation, Healthcare Triage, Network Optimization, Yield, Energy, and Throughput, R&D and Product Innovation, Procurement and Spend Analytics, Customer Service Optimization, Fraud and Debt Analytics, HR and People Analytics, Strategic Financial Analytics, and Accounting & IT.
Automation and AI are integral to the Fourth Industrial Revolution, driving innovation and productivity. The AI software market is experiencing rapid growth as a result, with a projected annual growth rate of 54%, according to Statista. Key applications include AI in E-commerce,
AI in the Auto Industry, AI in Agriculture, AI for Customer Service, AI in Logistics and Transportation, and AI in Surveillance.
A recent report from Stanford University underscores the rapid industrialisation of AI, emphasising its increasing accessibility and performance. In 2021, global AI private investments surged to $93.5 billion, which is more than double that of 2020, while the number of AI patents filed has grown 30 times since 2015. As AI transitions from consumer to industrial applications, it’s poised to drive digital transformation, spurring innovation, productivity growth, and advances in both theory and technology in the process.
By Prof. Mark Nasila, Chief Data and Analytics Officer in FNB Chief Risk Office