top of page

How to drive Artificial Intelligence in the workplace

Updated: Dec 14, 2022

AI can perform highly complex problem-solving (such as unravelling intricate cancer diagnoses), but it can also suffer major setbacks (such as the potential for racial discrimination) . One of the challenges facing organisations globally is the incorporation of AI in a way that provides people with the opportunity to succeed in the face of changing workplace conditions. This effort to understand how to harness AI’s capabilities offers insights into society’s historically and culturally configured fears, anxieties, and hopes about the future of work. The viability of AI as a technological alternative is therefore not only determined by the organisational forces that may resist or endorse it as such, but also shaped by human workers’ desire for such an alternative (Oxford AI Programme, 2022).

AI technologies are already generating value for organisations across different sectors including medical (diagnosis of disease) and financial (identification of fraudulent or suspicious financial activities) (Manyika & Sneader, 2018). Therefore, as AI technologies are increasingly used to complement human labour, organisational strategies will need to be adjusted to ensure that AI’s value can be fully utilised. AI presents promises in its potential to innovate in organisations (Oxford AI Programme, 2022).

Here is an example business case to implement AI technology using one of my client's strategic workforce planning business process.

#JungianBitsofInformation's framework to drive Artificial Intelligence in the workplace.


The SWP delivery board initiate the business case by agreeing a mandate to implement AI.

Problem Statement

SWP is a decision-making process that analyses current v future workforce and future pipeline of work data to predict the gap between supply v demand and develops action plans to close the gap. Data is integral to SWP decision-making. There is vast amounts of data, however, it is held in unconnected, disparate systems and software and limited resources to analyse the data. There is no data analysis capability: a labour-intensive, manual, time consuming, inefficient and costly task, and prone to errors that impede the board’s ability to make informed decisions in a timely manner.

Shared Vision

SWP is a priority in the People Strategy supporting the organisation's strategic objectives. SWP aims to anticipate the mission critical workforce required in 1-5 years, therefore, implementing AI in the process has the potential to contribute to the organisation's success. The board share the same goals and workforce challenges presenting an opportunity to create a shared vision for the specification and utilisation of an AI model.

Business Case

The use of AI in the SWP process is technologically feasible and economically viable. AI can outperform humans at narrowly defined, repetitive tasks which is the space in which AI excels, and where the main application of AI in the short/medium term can be found (Oxford AI Programme, 2022).

AI’s has the potential to bring immense value in its ability to analyse vast amounts of data and predictive capabilities which exceed human capabilities in terms of judgement and decision making. AI excels at recognising trends and patterns, forecasting, determining the relationship between variables and accurate predictions.

AI is best deployed on a well-defined process and repetitive tasks like data analysis. The right AI tool complements SWP decision-making. There are limitations to AI - it is difficult to automate decision-making. The AI model will be limited to predictions which augment rather than replace decision-making giving the organisation a competitive edge in the war for talent with more accurate diagnoses.

The type of ML model that can potentially solve the problem is Deep Learning and/or Supervised Learning which is known to produce accurate predictions based on independent variables i.e. current v future workforce, future pipeline of work.

There is currently no resource dedicated to data collection and analysis therefore the adoption of AI is unlikely to be met with resistance.

The estimated costs include:

  • Project management

  • Expertise in data and ML engineering

  • Integration of the AI model into existing systems using Application Programme Interface (APIs)

  • Operation costs

The economic case for the AI model is based on recovering implementation costs and on savings and efficiencies made as a result of accurate predictions which reduces our significant recruitment-to-hire costs and ensures the right workforce is in place at the right time to deliver the organisation’s strategic objectives.

The risks include:

  • Data is held in disparate, unconnected systems and software which may affect the accuracy and reliability of the data

  • DL’s hidden layers brings complexity in that it is difficult to understand how DL produces accurate predictions which can raise trust issues

  • The delivery board may over rely on the algorithm for decision making using the AI model to replace decision-making.


The ML model will be tested to ensure that it meets our needs before it is deployed. The algorithm will be tested using training data until the performance measures indicate it is working well. The algorithm will be deployed after it passes the acceptance criteria using an approach known as ML Operations: a set of practices that combines Machine Learning, DevOps and Data Engineering proven to deploy and maintain ML systems in production reliably and efficiently (Cristiano Breuel, ML Ops: Machine Learning as an Engineering Discipline. Towards Data Science, 2020).

Successful deployment depends on this combination of unique skills which together focus on the problem, automation and performance of the algorithm.

Operate & Improve

Post deployment will involve maintenance and continuous improvement to ensure the ongoing success of the ML model. The algorithm will be reviewed on a regular basis using performance indicators to determine when it needs to be retrained or to rectify errors. A set of guiding principles will be used to address any ethical or legal issues emerging from the predictions which may affect organisation's reputation.

In the coming months #JungianBitsofInformation will be offering a new service to organisations. An Artificial Intelligence service: to identify opportunities for AI in your organisation and guidance on the ethical considerations to address the common pitfalls of AI with a unique perspective from #analyticalpsychology. Register on the site and be the first to hear about the launch of this new service.


bottom of page