Liam Cotter charts the road ahead and critical importance of data for Irish organisations preparing for the AI revolution
Right now, many organisations are experiencing caution, confusion—or both—in relation to artificial intelligence (AI).
They are unsure about generative AI (GenAI), how it differs from previous AI iterations, and whether it can add value for them.
With the first milestones of the European Union’s AI Act due to come into force in February 2025, focused on prohibiting AI systems posing unacceptable risk, organisations are concerned about falling foul of regulation.
They are keen to ensure that any AI model introduced to help their business, undergoes rigorous testing to ensure it is fair and doesn’t have bias baked in.
There are also more generalised fears regarding the cost of moving too quickly and developing the wrong solutions, however, as well as the “opportunity cost” of moving too slowly and thus failing to capture the benefits of the right opportunities.
Data-based decisions
Regardless of what stage an organisation has reached in its adoption of AI and GenAI, one thing holds true: the key to success is data.
The only way to ensure quality AI outputs is to provide quality inputs. The way we manage and store data for the AI age differs from how we have done so in the past.
Thus, even though the same fundamental rules apply, your data capture and entry systems may not be robust enough to handle AI demands and this could put you at a competitive disadvantage.
Part of the problem with readying your data for AI transformation is the sheer amount of hard work involved, which may not appear not to offer a lot of value. This is because this work involves run-of-the-mill data generated from day-to-day operations.
The key to the successful adoption of AI tomorrow is ensuring everybody in your organisation is aware of data management today.
It is about ensuring everyone is measuring the quality of their data right across the organisation so they can stand over what it presents.
For organisations that previously placed little value on the data they generate, this shift will require a culture change. It may also require different parts of the organisation to pool data—such as combining sales and stock databases rather than keeping them siloed, for example.
In companies involved in mergers and acquisitions, it means ensuring you fully understand your data's lineage.
The time to act is now
The past 12 months have seen a growing realisation among organisations of the potential importance of AI as a lever for competitiveness.
It is increasingly viewed as a valuable tool to drive digital transformation, enabling them to become more flexible, be faster to market, provide a better customer experience and more.
Most of what AI will do has yet to be “dreamed up”. To put its scale in context, somewhere in the world, a data centre—the building block that powers the AI revolution—opens every two days.
Organisations need to act to keep up. The first step is understanding the regulations and timeframes that are being rolled out under the EU AI Act.
Next, identify use cases and develop them. Experiment—and if you are going to fail, fail fast. Get involved and discover the value in AI.
People-powered data
Understand the behavioural risks, too. A lot of the work involved isn’t about technology at all. It’s about people.
You can introduce the best technology in the world, but it's useless if staff don’t collect, curate and manage their data correctly.
Everyone in your organisation must be able to stand by the accuracy of their data, which means good data practices must be applied to all business processes.
In many organisations, this means investing in data capabilities, including staff training, and appointing a Chief Data Officer responsible for driving data literacy and good data management practices throughout the organisation, from the bottom to the top.
To succeed, data management must be seen as a core, valuable component of what everyone does, regardless of their role.
Break down the barriers
Barriers to achieving effective AI readiness include an organisational culture that hasn’t yet caught up with the importance of data, allied to poor systems and processes that ensure people don’t understand the implications of getting it wrong.
The real barrier is, however, that all of this takes work.
Readying your data systems for AI is a pain, and sometimes, people can see no value in it.
Once you can stand over your data, knowing it is of good quality and understanding its lineage, your organisation will likely be in pretty good shape because you can then move on and digitise your key business processes with confidence.
The AI revolution starts and ends with data. Don’t underestimate the effort required to get good quality, well-managed data. It is the foundational work that cannot be avoided.
Equally, don’t underestimate the impact. Once you have good data systems in place, you can confidently move forward and capture the full breadth of AI benefits that await.
Liam Cotter is Technology Practice Lead at KPMG
LOADING...