Brands should be cautious about how data is sourced in their AI models.
This piece argues that marketers should give careful thought to how consumers discuss their interest in a brand on social media, and how the data is sourced while training datasets or fine-tuning public AI models. It would help them identify inherent biases in AI models.
These inherent biases in AI models can make them less relevant to the market research industry. There is often a disconnect between insights gathered from different sources like traditional brand concepts shared by researchers and clients, and new data sources like social media.
The author suggests marketers should fine-tune their AI models to be sensitive towards specific contexts. Businesses can leverage “transfer learning”, a popular field of AI research and application, to modify their AI models.
[4 minute read]