Lightbulbs in a row

LWA/Getty Images

Companies might be keen to exploit artificial intelligence (AI), but research suggests that making the most of emerging technology is easier said than done.

As many as 87% of data leaders say AI is either only being used by a small minority of employees at their organization or not at all, according to Carruthers and Jackson’s Data Maturity Index.

Also: I’m taking AI image courses for free on Udemy with this little trick – and you can too

The consultancy’s annual poll of data leaders suggests many organizations are suffering from an “AI-induced paralysis”, where only 5% of businesses boast a high level of AI maturity, established AI departments, or clear AI processes.

However, it’s important that data leaders who feel their organization is lacking maturity don’t get too downhearted.

Caroline Carruthers, CEO at Carruthers and Jackson, told ZDNET that every new technology goes through a period of justification, governance, and acceptance.

“We’re all on a journey,” she says. “We’ve got a lot more data than we’ve ever had before. Data is fundamental to our businesses.”

As a starting point for building AI momentum in slower-moving organizations, Carruthers suggests four priorities for data leaders who want to move beyond the current AI paralysis:

  1. Start with purpose — “I can’t stress that enough. What do you want to do? What problem are you trying to solve? What keeps you awake at night? What opportunities do you have? What excites you? You need to have some reason to move forward. And without it, we’re looking like a bunch of kids playing sports on a Sunday. We’re just scattered all over the place. So first and foremost, concentrate on purpose.”
  2. Focus on targeted outcomes — “What’s the smallest part of that purpose that you can start making a difference on? When you go down this path, and as soon as you mention things like AI, everybody goes for ‘bigger is better.’ It’s like, ‘What’s the biggest problem? Can we solve world peace?’ Instead, focus on the smallest problem where you can make a difference and use that as your model going forward.”
  3. Shout about your successes — “Data people aren’t very good at telling everyone about the good stuff they’re doing. We’re very good at thinking about how much we have left to do. And we’re very good at running around and doing lots of stuff. But we’re not very good at going, ‘Look at this great stuff we have,’ and encouraging people to come on the journey with us.”
  4. Use data to prove your case — “Show people the results of your project. Did it work? Did the AI do the stuff that we told everybody it would be doing? Could we have completed the project better or faster? Understand the metrics, so that you can get buy-in for more projects.”

Focusing on those four priorities will help your organization to start building AI momentum.

But given all the hype and excitement for generative tools, such as OpenAI’s ChatGPT and Microsoft Copilot, why does AI remain at such a nascent level of development?

Carruthers says the explanation is simple — embracing AI involves an ability to overcome two big hurdles: people and regulations.

Hurdle 1: The people problem

When it comes to people, all kinds of employees in the business — from the boardroom to the shop floor — need to be convinced of the value of AI.

And Carruthers, who is a former chief data officer (CDO) of UK infrastructure giant Network Rail, says convincing people is no easy task, despite all the excitement surrounding the rapid growth of generative technologies.

Also: Why open-source generative AI models are still a step behind GPT-4

“As soon as you mention the word AI, people visualize Skynet and start thinking they’re going to lose their jobs,” she says, referring to both the fictitious AI system in The Terminator and the very real fear that many people have about the potential impact of emerging technology on workforce numbers.

“While many data leaders feel they need to be doing something with AI, they also face an intrinsic level of resistance built-in before they can even start doing anything.”

Hurdle 2: The regulatory bind

When it comes to regulations, the Carruthers and Jackson research suggests executives are rightly concerned about data ethics and the potential for more stringent data laws focused on the use of information.

However, as the form of these rules and laws is still unclear, many companies are choosing to bide their time before pushing headfirst into AI.

Also: The ethics of generative AI: How we can harness this powerful technology

“It’s a bit like smoke and mirrors. Legislation is coming — we know lots of people are talking about it, but we don’t know what those laws mean yet,” says Carruthers.

“So, I think people are hedging their bets a little bit because they don’t quite know what’s going to come.”

Momentum needs solid foundations

The research suggests the tricky combination of a fearful workforce and the unpredictability of the current regulatory environment means many organizations are still stuck at the AI starting gate.

As a result, not only are pilot projects thin on the ground, but so are the basic foundations — in terms of both data frameworks and strategies — upon which these initiatives are created.

About two-fifths (41%) of data leaders said they have little or no data governance framework, which is just a percentage higher than the previous year’s Maturity Index, when 40% of data leaders said they have little or no data governance framework, which is a set of standards and guidelines that enable organizations to manage their data effectively.

Also: IBM says generative AI can help automate business actions

Just over a quarter of data leaders (27%) said their organization has no data strategy at all, which is only a slight improvement on the previous year’s figure (29%).

“I get why not everybody’s quite there yet,” says Carruthers, who, as a former CDO, understands the complexities involved in strategy and governance.

Moving to your data-led goals

Carruthers and Jackson’s research suggests the key role of governance means companies that want to be ready to exploit AI must focus on the creation of a data strategy and a supporting data framework.

“We have to put something in place that we haven’t had in place before to understand the implications of what AI can do and the good that it can cause,” says Carruthers.

Also: Generative AI fails in this very common ability of human thought

The good news is some digital leaders are making headway. Andy Moore, CDO at Bentley Motors, is focused on building the foundations for the exploitation of emerging technologies, such as AI.

He explained to ZDNET recently how he’s created an enterprise-wide data strategy around four core pillars: governance; the data cloud, which is Bentley’s technology stack; the data dojo, which is his internal data literacy program; and enablement, which focuses on helping the data team work with the rest of the business.

Also: ChatGPT Enterprise vs. ChatGPT Team: Which is the best for your business?

“My ongoing challenge as a data leader is to set the possibilities for these technologies without saying you can have it now — because, of course, everybody wants AI now,” he says.

“I need to say, ‘I can’t give you AI now because I’ve got to get the foundations in place first.’ So, my role is about balancing expectations, while still moving at the pace the business needs as well.”


Source link