In this session, Beata Kouchnir, Director of Machine Learning Science at Glassdoor and Anna Bethke, Ethical AI Data Scientist at Salesforce, discuss why human-in-the-loop and high-quality data go hand-in-hand to achieve widespread production AI applications.
Here are 3 key takeaways from the session:
- With the proliferation of production AI, enterprises realize that when machine learning can’t solve the problem, humans need to intervene. To account for this gap, enterprises are integrating robust human-in-the-loop programs both for training the model and evaluating the performance of the model.
- Given the state of AI today, humans-in-the-loop will always play a part. The role may only change from pre-training to a more complex post-production role like monitoring and drift detection.
- Automation is good for dull, dirty, and dangerous tasks. Tasks that require reasoning, cognitive processing, and creativity would require human intelligence for better performance.