On 21st July, at the esteemed MachineCon 2023 held in New York, Krishna Gade, the Founder and CEO of Fiddler AI, delved deep into the intersection of Generative AI and Responsible AI. His insights offer a comprehensive understanding of the challenges and solutions associated with operationalizing AI models, emphasizing the importance of fairness, accuracy, explainability, robustness, and privacy.
Building Trustworthy AI Systems
Krishna began by posing critical questions: How can we ensure AI models are developed with fairness, accuracy, and robustness? How do we manage and monitor these models once they’re in production? He stressed that in high-stakes domains like financial services, healthcare, and manufacturing, trust is paramount. To build this trust, there’s a need for rigorous model validation, continuous monitoring, and robust governance.
The Risks and Challenges of Operationalizing Generative AI
Operationalizing Large Language Models (LLMs) and other Generative AI applications come with inherent risks. Krishna highlighted the challenges stakeholders face when bringing AI/ML models into operation, especially from a human-centric perspective. He emphasized the importance of adopting responsible AI practices throughout the model’s lifecycle, from validation to post-deployment monitoring.
Tools and Techniques for Monitoring Deployed ML Models
Krishna provided an overview of the tools and techniques essential for monitoring ML models once they’re deployed. He shared industry-specific cases, shedding light on the practical implications and benefits of these tools.
About Krishna Gade
Krishna Gade is not just the founder and CEO of Fiddler AI, an AI Observability startup, but also a seasoned engineering leader with a rich history of creating scalable platforms and products. His experience spans across tech giants like Facebook, Pinterest, Twitter, and Microsoft. At Facebook, Krishna was at the helm of the News Feed Ranking Platform, which powered various features, including Facebook Stories and recommendations. He was instrumental in introducing Facebook’s explainability features, such as ‘Why am I seeing this?’, bringing transparency and accountability to the platform’s algorithms.
In Conclusion
Krishna Gade’s talk at MachineCon 2023 was a clarion call for the AI industry. As we venture deeper into the world of Generative AI, it’s crucial to ensure that the systems we build are not just innovative but also responsible. By emphasizing the importance of transparency, fairness, and robust monitoring, Krishna has set the tone for the future of AI.
