AI Fundamentals

As can be seen from the diagram above, AI is simply 1 component of Digital Business Transformation. There seems to be a lot of confusion regarding the 4IR. Breakthroughs in emerging technologies in fields such as robotics, artificial intelligence, nanotechnology, quantum computing, biotechnology, the internet of things (IoT), the industrial internet of things, decentralized consensus, fifth-generation wireless technologies, 3D printing, fully autonomous vehicles, digitization, blockchain, virtual reality, augmented reality and so on.

It is critical that young people and business in general have a really robust understanding of this new technology. SCM and Logistics, finance function, HR function, production and operations as well as banking will change business processes dramatically. It is not coming, it is already here, so do not fall behind. Students must ensure before enrolling for a B.Com, for example at a particular University must ensure that DBT is part of the curriculum, right up to third year, otherwise the degree will be redundant very soon.

If we look at the transformation process, we see inputs - transformation process - outputs and feedback loops. The inputs are labour, land, capital, labour, buildings, equipment, technology and so on. Today we can add cloud services and artificial intelligence as "factors of production" in the input stage. Remember that AI depends upon cloud-based computing power, the development of algorithms, and mountains of data (big data). AI is in fact the study of computational mechanisms underlying thought and of course intelligent behaviour, and AI is based on two fundamental technological capabilities: human perception and human cognition. There are three technological advances that have provided the launchpad that has allowed AI to take off. Firstly, computing power has finally advanced to the level needed to perform the enormous number of calculations needed. Secondly, cloud computing has made large amounts of this power and storage capacity available to both people and organisations without the need to make large capital investments in massive amounts of hardware. Thirdly, the explosion of digital data made it possible to use massively larger data sets to build and train AI models.

"Unsupervised learning" involves the use of machine learning to train an AI model with unsupervised data, and the use of unsupervised learning systems is making access to larger amounts of more advanced computing power more important than ever. Facial- recognition technology, using AI-based computer vision with cameras and data in the cloud, can identify the faces of customers as they walk into a store based on their last visit. I believe that this should be regulated due to matters of privacy. The use of unsupervised learning systems needs to be coupled with access to data - data about the physical world, the economy, and in essence how we live our daily lives. Application programming interface (API) enables third parties to use GPT-3, which is a third generation natural language prediction model.

If data came from individuals, legal issues around privacy will need to be worked through (regulated). And if the data does not involve personal information, other big questions need to be hammered out, such as governance processes among organisations and the ownership in data as it grows and is improved. These challenges are not just technical in nature, they are also organisational, legal, social, and even cultural.

Regarding new legislation, we can use a concept that has been championed for start-up companies and software development, referred to as a "minimum viable product". As defined by Eric Ries, it advocates creating "an early version of a new product that allows a team to collect the maximum amount of validated learning about customers". AI naturally gravitates toward monopolies, so that once a company has jumped out to an early lead, this kind of ongoing repeating cycle can turn that lead into an insurmountable barrier to entry for other firms. This is referred to as "network effects". So data could be locked up and processed by a few giant tech companies, and every other economic sector will rely on these tech companies for their AI services. As machine learning has evolved rapidly over the past decade, it has become apparent that there is no such thing as too much data for an AI developer.

The importance of computer science and digital skills has being driven by two primary long-term factors. The first is the emergence of AI-powered technologies that are propelling a new era of automation. The second is the growing need for firms to compete more effectively in the changing commercial landscape. All these challenges require action. The first prize is to do something big. Second prize is to do something. Success rarely comes to people who do nothing.

Prof Rory Dunn.