Emerging technologies promise big benefits in theory, but can be hard to deliver in practice. Research from the Capgemini Institute shows that the adoption of generative artificial intelligence (GenAI) is still in its early stages, with nine in 10 companies yet to scale up these early projects.
But boards of directors are increasingly pressuring CIOs and their teams to gain competitive advantage through innovation. So what are the key things business leaders have learned about AI so far? Four business leaders share their tips.
And it's time for companies to look beyond the hype around generative AI and find real value.
1. Always involve humans
Miguel Morgado, senior product owner for Eutelsat Group's Performance Hub, said the company's use of AI and machine learning relates to predicting outages and analyzing their root causes, such as how weather can affect satellite dishes.
The exploration of these emerging technologies demonstrates the importance of high-quality information.
“We do a lot of testing with real data,” he said. “And validating the model is really important. If you don't have an accurate model, then using it, you're in a 'garbage in, garbage out' situation. Having a good dataset is important.”
Morgado said his company is fortunate in that the satellite company collects billions of rows of data every day for a variety of uses, but the company ensures that this information is applied safely and effectively.
“The models and results will need to be tested multiple times until this approach proves correct,” he said. “The results will not be perfect — there will always be some imperfections — but it's an indication.”
And when is the right time to invest in AI? 4 ways to help you decide
Morgado told ZDNET that other companies should ensure that trained professionals are always on top of things and communicate the importance of results to colleagues.
“That way, you can have that person say, 'This result shows a certain value, or it could be taken as guidance,'” he said. “So ultimately it's always the user who decides whether to trust the AI or not. My advice to others is to always make sure there's a human element in your AI results.”
2. Get senior buy-in for organizational change
Ulf Holmström, chief data scientist at Scania Group, said the company is looking at how to use AI for its internal support processes.
The company: Amazon Bedrock And they're eager to explore how they can leverage Snowflake's tools. Cortex AI.
Like other business leaders, Holmstrom pointed to the importance of underlying data and technology concerns.
“Whatever you call it, you need to have confidence in your data and your infrastructure and your governance. Otherwise you can't scale up and you can only do proofs of concept. Like any organization, we need to get our products into production.”
Plus: Agile development can unlock the power of generative AI – here’s how
Holmstrom told ZDNET that the good news is that technology is becoming easier and easier to adopt, as access to technical knowledge is being democratized through technologies like the cloud and generative AI.
But Holmstrom said new processes would need to be put in place to allow users to take full advantage of the emerging technology.
“When you introduce AI into manufacturing, there are implications. One of the big implications is that you have to change the way you work. That means new business processes and new types of organisation,” he said.
“We will need to learn new skills and change the way we work. That change will be difficult for all organizations, but especially for legacy companies. But without that transformation, AI will never be successful.”
Holmstrom said senior management needs to drive the transition to this new way of working. “Commitment from top management is crucial,” he said. “AI transformation can never happen bottom up. It has to happen top down.”
3. Remember that real-world biases exist
Anastasia Stefanska, data analyst for analytics and AI at travel company TUI, said it was important to think about how to turn the vast amounts of data collected into high-quality information, and that part of that work required being aware of human bias.
Like other experts, Stefanska recognized that ensuring organizations have high-quality data is a prerequisite for successful AI projects.
But this is not the only important issue: smart practitioners will also consider real-world biases while focusing on data quality concerns.
“AI is a simplified reflection of the realities of the real world we live in,” she says. “But it's paramount that we drive our adoption of AI with data quality in mind. But we can go beyond data quality by looking critically at the current state of the real world, and think about opportunities to address biases that are deeply rooted in the real world.”
And four ways organizations can overcome AI inertia
Stefanska spoke to ZDNET about how TUI is using the Snowflake platform to consolidate corporate information and build a digital platform for data-driven change.
As part of this work, Stefanska and her colleagues monitor how the data is used and misused.
“That's why at TUI we say the human eye matters. We accept that biases exist in the real world around us,” she said.
“My main message is: yes, data quality is important, but take a holistic view of the amount of data you have and whether there are opportunities to transform it into something of new quality.”
4. Use AI when it makes sense for your business
Richard Wazatt, CEO of foreign exchange specialist Travelex, advised other experts not to walk before they run. He acknowledges that there is a lot of hype about AI, but fear of being left behind should not dictate their judgment.
“We're not looking to be an early adopter of AI right now,” he said, “but as the case for how it has helped other companies becomes more proven, we will embrace it.”
Wazacz told ZDNET that his extensive business experience as a director at Octopus Energy has helped him develop a strong appreciation for an era in which emerging technologies will play a key role.
“I worked at Octopus and they've done a good job of using AI to improve customer service,” he said. “A lot of customer questions are answered through AI. Do you think we now have the option to do that too? Yes, they've proven it's possible, so it's less risky.”
Also, do AI tools make it easier to start a new business? 5 factors to consider
Wazatt says that Travelex's approach is: Mesh AITo ensure that digital investments are made in the right places.
Mesh-AI has helped Travelex establish a cloud-based data platform, initially focusing on real-time reporting, and the company plans to expand into other emerging areas when the time is right.
“I'm staying very narrow,” Waszak says, “and that's exactly what Mesh-AI is doing right now. They're excited about what they're doing. They're taking the approach that, 'If we prove to our customers that they can stand on their own, we'll get more business,' and I feel like they're doing just that.”
