VIDEO | UST UK Head of Data: AI Is Nowhere Near Sophisticated Human Brains

(US and Canada) Heather Dawe, UST UK Head of Data, speaks with Kellie de Leon, Treasure Data Senior Director of Content Marketing, in a video interview about the challenges, opportunities, and risks of using AI.

Dawe points out that the challenges of applying AI are multifaceted and says that, while AI can automate repetitive tasks and do other good things, it is nowhere near sophisticated human brains. Similarly, while it can be more consistent than humans, it does not understand nuance and sophisticated arguments yet, she says.

Sharing an example of sophisticated and advanced usage of artificial intelligence, Dawe mentions combining the expert knowledge of cardiac surgeons with advanced analytics to create augmented intelligence. However, she maintains that the industry is still at a relatively basic level with AI usage.

When asked about disconnects between AI and ML among business leaders, Dawe says that one of the most common misunderstandings is that AI can solve all problems. She explains that, for AI to be effective, robust data sources are needed to underpin the analytics, which is connected to data management and literacy standards within organizations. Similarly, machine learning is not a silver bullet, she adds.

Speaking on ways to operationalize data in the business, Dawe suggests that it is not about technology but the cultural changes that are required for an organization to be truly data driven. She believes the CDO brings in a lot of promise as they have to be data evangelists. Dawe further points out that large businesses have huge amounts of reports and KPIs, but they have to be consistent and accurate in reporting.

Discussing the risks of using AI and ML, Dawe indicates that there are a lot of things to be aware of to develop models, make them live, and ensure that they are ethical and fit. She states that society, in general, is biased, which means biased data gets used in machine learning models, and the outcomes will carry bias as well.

Dawe concludes that the awareness of bias impact and ensuring that the AI is fair and fit are key.

CDO Magazine thanks Heather Dawe for contributing her thought leadership to our global community.

See more from Heather Dawe

 It's Still a Challenge to Deploy Machine Learning Models at Scale