The potential of Generative AI to be a transformational enabler in future business and operating models is becoming increasingly apparent, as is the importance in starting things off with the right footings, before attempting to scale writes Owen Lewis, head of AI at KPMG Ireland.
There are a few considerations that are important. First is getting the basics right in terms of ethics, governance, data foundations and above all - delivering measurable and tangible value.
To harness the full potential of Gen AI, organisations must think about getting their data foundations right. This involves understanding what data you have, what it was intended to be used for, what you want to do with it, and importantly what you ethically should do with it. Organisations are establishing data governance structures to deal with these types of questions both to be compliant with regulations, and also to ensure they are embedding AI enablement within the strategies of their business to deliver real value.
AI has long been embedded in the tools we use and is already hugely valuable. But while we know GenAI has enormous potential value for enterprise, many businesses are still in the midst of actively figuring it out.
Most organisations now understand that GenAI, at least at this point in maturity, is there to augment and needs an expert to ‘sit with’ it – the human in the loop. But we are seeing hype-versus-reality is now leading to an expectation risk, and some unease in where to get started, and make the most of the newfound capability at our fingertips.
Expecting GenAI to perform a magic trick is not going to be helpful. It’s important to appreciate the need to spend time explaining to a GenAI platform what you are trying to achieve, the information you want it to use, and the output you want it to create. Just as you would with a colleague.
Simply asking GenAI to perform a task with little to no context and being disappointed when it doesn’t do a very good job, is a potential risk for the technology in the short term too, as people get dismissive that we are not seeing valuable outputs.
The reality is that you have to work at GenAI, and the human has to play a role both on the input side and then in the quality assurance of the output.
Use cases growing
What is exciting is that valuable applications or use cases are being discovered all the time. A central pillar being boosting personal productivity. Depending on the user’s role, this can be anything from scheduling and preparing for meetings to post-meeting analysis. GenAI is also speeding up research and beneficial when generating the first draft of documents.
One of the earliest measurable use cases has been seen in the software community, particularly those who write code and develop test cases. Previously, a software engineer might spend 30 per cent of their time deciding what they want the code to do, and 70 per cent then writing it. But AI- powered tools have flipped that ratio. It has enabled software engineers to become engineers in the real sense, focusing on architecture and how things fit together, and getting help from AI to turn these ideas into reality. The result is higher performance all round. Of course, you still need human capability to drive it, but GenAI is helping yield enormous productivity benefits.
GenAI is being put to work optimising solutions for everything from climate change innovation, to supply chain optimisation, and from energy grid use to bus scheduling, while at the same time generating process documentation required along the way, ensuring systems are much more streamlined overall.
The next breakthrough is the intelligent agent, combining automation and AI to operate across multiple process steps, and orchestrating various tools we use as humans. This will see a movement from the current question and answer mode seen with many GenAI interactions, to objective and real actions. This is where productivity will significantly increase.
Risk controls
Going about this in the right way is critically important to build trust in AI platforms. With the advent of the EU AI Act, the first elements of which come into force in February, some organisations might be reluctant to embrace AI for fear of falling foul of the legislation. Yet what the legislation actually does is make clear what AI can and can’t be permitted to do, clarifying the situation and reducing that risk enormously. Embracing this is important and really helps businesses set out foundations so they can scale with confidence.
Similarly, ensuring compliance with other data-centric regulations such as GDPR are critical to ensure that what you use AI for, and how you use it, means you can safely drive value from your data.
The road ahead
For organisations at the foothills of GenAI, the first step is to be aware of both the opportunities and risks.
Consider how you want AI to be thought of in your organisation and with your clients and customers, and the governance you need to put in place to ensure you are using AI tools in a trusted way.
Then figure out the strategy you are trying to achieve and how AI will help you deliver tangible value. Understand the fundamentals you need to put in place to make that happen.
Build momentum through real use cases, education, and demonstrating real value - and scale this across your workforce.
If you are clear on governance, clear on strategy, and have considered the fundamentals and how you are going to manage them, then the opportunities are endless.
Get in touch
At KPMG we understand the pressure business leaders are under to get it right on tech and AI.
To find out more about how KPMG perspectives and fresh thinking can help your business please contact Owen Lewis of our AI team. We’d be delighted to hear from you.