Prerequisite: None
The abrupt emergence of AI as a practical enterprise utility has inspired organizations to rapidly prototype and integrate AI into a broad array of applications. Yet in the excitement and hype of deploying the technological componentry for AI, organizations tend to overlook two aspects of information quality that are critical to AI success: the quality of the data used to train and fine-tune AI models, and the quality of the information produced by AI systems that rely on transformer or generative algorithms.
In this session, TDWI Research Fellow David Loshin will discuss the concept of information risk in the context of enterprise AI deployment. He will then outline how the data and information quality technology market needs to transform to be able to address information risk concerns.
Attendees will learn about:
- Generative model dependency on massive volumes of information
- Inherent information biases that can infect your models
- The risks of AI confabulation
- How automation bias can catch you off guard
- The needs for continuous verification and validation