ChatGPT and what advanced AI means for operations

Businesses have long been trying to figure what the future of artificial intelligence (AI) and machine learning (ML) might mean for them. Following the emergence of OpenAI’s latest version of ChatGPT, the conversation has been accelerated: advanced AI is no longer theoretical, business leaders need to understand its potential threats and opportunities now.

How is ChatGPT changing the conversation on AI?

There are several reasons organizations might use ChatGPT. Number one, it can analyze extensive amounts of data and then provide predictive analytics, which can potentially enable smarter decision-making. This is the most relevant capability for ops – the ability to translate data into business insights and then into possible decision proposals, saving time and effort that would otherwise be spent trying to make sense of the data. The second reason is process automation – ChatGPT can learn about the process based on KPIs and data, enabling managers to make decisions and simplify the process cost. For example, more advanced chatbots can better understand and resolve questions. But the biggest area where we have seen the use of ChatGPT increase is to improve customer services and customer engagement. It is very easy for ChatGPT to personalize services for customers, such as providing targeted recommendations based on a customer’s specific needs.

When it comes to operations, there are many ways ChatGPT and other advanced AI tools could be applied. For example, AI can look at team structures and give a view of the ideal team-lead-to-team-member ratio based on workloads. It can look at the skills of individuals and recommend work based on their performance across different tasks. And it can look at a range of performance indicators and then provide predictive analytics on the back of that, for instance, whether certain employees are more productive if they are working from home or in the office. Ops managers can then either act on those insights or choose to ignore them, but those predictive analytics can help drive efficiency gains across the operations function.

The drawbacks

There has been some pushback against ChatGPT. Some of the largest global investment banks have either completely removed access to ChatGPT or restricted its use. That is partly driven by risk management concerns – when you transfer data from the bank to ChatGPT to generate insights, there are issues around data privacy, security and ethics.

In addition, ChatGPT isn't always accurate, meaning organizations must check the veracity of any information provided, which can be a challenge if it is unclear where that information was sourced. There is also growing uneasiness about AI becoming more intelligent than humans – a fear even shared by AI ‘godfather’ Geoffrey Hinton.

Some companies have faced criticism for how they are using AI for monitoring employees. AI-driven analytics on employee demographics, skills, and performance data could be used to support progression – or misused as a source of bias. The lesson here is that when using algorithms, organizations must build in human empathy and morality otherwise the AI will just optimize based on the numbers presented.

What the future holds

Like all innovations, adoption is unlikely to happen overnight. Financial institutions are likely to take a phased approach to the use of ChatGPT and AI tools. Some aspects might see faster adoption, such as making the customer experience more personalized. For automation, it might be longer depending on what progress those organizations have made with their digital transformation efforts and how easily they can access data to feed into the AI tool.

Ultimately, advanced AI tools like ChatGPT have significant potential to shake up how operations are run, given that AI can analyze vast quantities of data and propose directions far quicker than humans ever could, and in an objective way. How soon that will be remains an open question.