AI Is a Tool, Not a Replacement: DataNoble CIO
Tiaan van Zyl on AI Limitations and Human Accuracy in Customer Feedback
DataNoble, a customer experience management firm, uses telephonic and video interviews to collect feedback on customer journeys, SWOT analyses and KPIs. These interviews are recorded and manually transcribed, with responses typed into a form.
"For accuracy, we do everything manually. Transcribing using voice-to-text is problematic because it makes mistakes. We want to represent what our respondents say accurately, and we also wish to convey the conversations that we have in a way that is relevant to the questions we posed to them," said Tiaan van Zyl, CIO of DataNoble. "We've got a lot of digital tools but the human element is extremely important. These are things that machines don't do that well yet."
The company looks "at a variety of factors" when considering new technologies.
"Reviewing the interviews gives us an understanding of what the person said. We cannot rely exclusively on AI to interpret these things; we use machines as tools. We don't use machines as replacements [for people]," van Zyl said.
In this video interview with Information Security Media Group, van Zyl also discussed:
- Why humans excel in accuracy over machines for specific tasks;
- Balancing AI automation with human insight in data analytics;
- A CIO's perspective on the CrowdStrike outage.
van Zyl started his career as a researcher at a skills development organization in Pretoria, South Africa. His first role in IT was as a software developer for the South African Institute of Stockbrokers, where he developed their membership and education systems. Over the past decade at DataNoble, he has spearheaded the organization's digital transformation. van Zyl works closely with the executive teams of DataNoble’s customer organizations to ensure that their research programs are aligned with their goals.
