Dr. Scott Zoldi, the Chief Analytics Officer at FICO, has been working in the field of artificial intelligence for over 25 years. He believes that mistakes are being made due to a lack of critical questioning and understanding when it comes to machine-learning models.
Zoldi is concerned that new adopters of AI are using models without being able to explain how they work, leading to the introduction or perpetuation of biases that can harm customers. To address these concerns, he has recently submitted patents that focus on transparency in AI model development.
The first patent, which has been granted, involves codifying AI model development in a blockchain to create an immutable record of the work done. This record shows when mistakes were made, rejected and remediated providing much-needed transparency. The second patent, which is pending, focuses on ensuring that the person who built the model documents how it was done, allowing for monitoring and understanding for years to come. Zoldi believes that transparency is crucial in the field of AI to prevent biases and mistakes.
In addition to his work in AI, Zoldi is also focused on educating future generations of experts in the field. He was the executive sponsor of the first FICO Educational Analytics Challenge which partners with historically Black colleges and universities to train students to work with complex data. The challenge focuses on identifying and mitigating biases in data with plans to tackle payment fraud data in the future.
Zoldi’s experience and expertise make him a valuable resource for those looking to navigate the complex world of artificial intelligence. He encourages others in industry not to get too comfortable with AI as he believes that mistakes are being made due to lack of critical questioning and understanding when it comes to machine learning models.
As more organizations continue to adopt AI technology, it’s important for them to consider Zoldi’s advice and prioritize transparency in their model development processes. By doing so, they can help prevent biases and mistakes from harming their customers or stakeholders.
Overall, Zoldi’s work highlights the importance of transparency and critical thinking when working with machine learning models
