Artificial Intelligence imperative for competitiveness
Artificial intelligence (AI) and Machine Learning (ML) have been with us for more than a decade. But it was only in the last few years that AI technologies started to advance at an unprecedented pace and created numerous new opportunities for businesses to redefine their operations and strategies.
According to a recently published letter by the well-known Bill Gates, “The Age of AI has begun“, he emphasises that AI is just as revolutionary as the PC, the Internet and mobile phones. Numerous areas will be transformed by AI, such as healthcare, education, and the workforce. The proliferation of AI will allow industries and businesses to become significantly more efficient.
Generative AI
One of the latest innovations in AI is generative AI – a branch of AI that focuses on creating or generating new and original content. It involves the use of machine learning algorithms, particularly generative models, to produce outputs that resemble human-created data such as natural language texts (stories, news articles, marketing material), images, music, videos, or even entire virtual environments (virtual worlds, virtual reality experiences) in response to prompts and user input.
Generative AI is capable of learning patterns and structures from large amounts of training data (machine learning) and then generate novel outputs based on that learned knowledge. It thus has the ability to automate content creation and drive innovation in various fields such as entertainment, design, marketing, and research.
While previous iterations of AI required human guidance, machine learning is able to independently observe data patterns with very little or no human interventions to make decisions and suggestions. Due to these power capabilities, generative AI has the potential to transform businesses and entire industries.
The chatbot, ChatGPT by OpenAI, which became hugely popular and accrued millions of users in a few months, is one of the renowned examples of generative AI and its self-learning capabilities.
Generative AI in business
Many businesses already adopted generative AI to reduce costs and increase efficiencies. Some examples are:
Operations: Making operations more efficient and increasing productivity by generating and executing task lists, as well as detecting production defects.
Marketing and sales: Creating and personalising marketing content; analysing and summarising customer feedback; providing sales support to the customer via chatbots to enhance customer experience.
Information Technology: Automating coding processes, auto-completing data tables, and AI code creation.
Translation: In the media world, generative AI is used for translation and content localisation in order to diversify across languages and regions. AI can translate a breaking news story in less than three minutes. The access to multiple regions obviously increased page views and ad revenue.
Predictive AI
Predictive AI, also known as predictive analytics or machine learning, is a branch of AI that uses algorithms and statistical models to analyse historical data and make predictions or forecasts about future events or outcomes. It aims to uncover patterns, relationships, and trends within the data to predict what is likely to happen in the future. These predictions can be used to inform decision-making processes, optimise business operations, improve customer experiences, and much more.
Predictive AI in business
Predictive AI is primarily used to automate human processes at very high levels of accuracy and with minimal human oversight. Predictive AI, therefore, has the potential to considerably impact businesses and its labour force in the future.
Some of the areas in business where predictive AI has made a significant difference are:
Cybersecurity: Due to the increasing cybersecurity threat, predictive analytics can automate the process of the monitoring and identification of potential threats, the detection of anomalies and patterns that may indicate a breach, as well as reacting to security incidents in real time.
Risk management: Predictive AI is very useful to assist with risk management, such as credit risk management and the prediction of the defaulting of customers.
Marketing and sales: AI is widely used by marketing and sales in business-to-business (B2B) and business-to-consumer (B2C) platforms. AI can sift much faster through a larger set of customer relationship management (CRM) data than a human being, thus increasing efficiency in marketing. Predictive AI can generate high-quality leads and assist sales in the conversion of leads into paying customers. On a B2C level, AI enables advertisers to more effectively segment populations into targeted groups to get the best return on their advertising.
Data-centric AI
From the above examples, it is evident that the quality of the data used by AI is just as important as the code and algorithms. Ensuring high-quality data that is consistently labelled would unlock the value of AI in business, manufacturing, and government.
Until now, data has often been collected and engineered in an ad-hoc way, which often depends on the availability and skills of individual data scientists and analysts.
Some of the problems with data that are frequently encountered are:
Inconsistent labelling: In some fields like manufacturing and pharmaceutics, AI systems are trained to recognise product defects. The same is true for the inconsistent coding of electronic records. AI systems should be trained on consistent data.
Emphasis on big data: The popular belief is often that more data is better. But in some cases, such as manufacturing and healthcare, large amounts of data is just not available. The emphasis should fall on the quality rather than the quantity of data. A typical example from healthcare is that there might not be many MRI scans of a given medical condition, or a factory might have only made 50 defective cars.
Ad-hoc data curation: Any data analysts knows that data can often be messy and has errors. In the past, this was solved by the skills of data analysts, but fortunately, most of the more advanced data analysis tools now assist with this important task.
Any AI system and its decisions are just as good as the data that is fed to them. This is the surest way of unlocking the value of AI.
However, in addition to data of high quality, standardisation is just as important. A ten-year-old MRI machine will generate different entries than a brand-new and improved one. Unfortunately, this is often a given and difficult to change.
The problem is that heterogeneity in the physical environment leads to a very fundamental heterogeneity in the data. If standardisation is not possible, these heterogeneity in data will need different customised AI systems, which obviously will have a cost implication for a business with different infrastructure in manufacturing plants or company branches.
AI and competitiveness
Artificial intelligence (AI) in business is quickly developing into a popular tool for competition. Companies across sectors will therefore need a clear strategy to incorporate AI if they want to remain a market leader in five years.
The need for Chief Artificial Intelligence Officers, data analysts, coders, and additional resources dedicated to this innovation is increasing as companies look to harness this innovation effectively.
The incorporation of AI within enterprises and wide-scale adoption by consumers will have a huge impact on businesses in the coming months and years. Companies that are harnessing generative AI to enhance their business and leveraging predictive AI to make their business smarter are set to disproportionately benefit from AI.
As companies embrace the AI revolution for businesses, they stand to gain significantly in terms of efficiency, cost savings, and new revenue streams.
Professor Louis C H Fourie is an Extraordinary Professor in Information Systems University of the Western Cape
BUSINESS REPORT