Navigating the AI transformation journey

The genealogy of AI as a professional discipline can be traced back to the 1956 Dartmouth conference in the United States where the phrase is said to have been coined. Alan Turing is often credited as one of the first theorists to ask if a machine can think.

Today, machines thinking is no longer a distant concept. AI currently requires human developed software and hardware, and commands, to execute tasks. It is possible that in the near future we will accomplish true AI, known as ‘General AI’ – i.e. computers performing tasks without human intelligence.

The other field of AI is known as ‘Narrow AI’, of which there are a number of subsets; including machine learning (training data consisting of an input and an output) and deep learning (deep graphs with many processing layers). Machine learning is often split into three forms: supervised learning, unsupervised learning (resulting in what is known as a ‘black box’) and reinforcement learning.

AI solutions are dependent on access to large and diverse datasets. These datasets are required to shape, train and direct AI towards the right outcomes. A model is trained through combining the weighted parameter (i.e. cleansed data that is assigned weights) with an algorithm to produce an AI data output.

Use of AI has a number of benefits, due to its ability to undertake high-volume and repetitive tasks in a shorter timeframe than a human. These benefits include enhancement of product or service features and functions, optimisation of business operations and processes, the ability to pursue new markets and the automation of business processes.

However, with advantages come a number of risks and challenges. From a commercial perspective, implementing and/or investing in AI affects an organisation’s business strategy, supply chain, customer relations and people. Ultimately, AI will impact society as a whole. This guide sets out relevant legal and strategic considerations for organisations embarking on this transformation journey.

Content last updated: 20 November 2020