LLM - Answer data sources
When designing AI-based products trust and transparency are paramount to drive adoption. Since deep learning algorithms like LLMs (Large Language Models) operate differently than traditional computing systems (they’re capable of learning and their output can’t always be exactly predicted), establishing trust is key in creating a successful AI product. Here’s how we’ve made sure Seam’s user experience follows key principles of designing trustworthy AI products:
Transparency - We make sure to clearly differentiate AI features from traditional ones. This way user’s expectations are managed and potential anxiety behind the model’s decision making is alleviated,
Continuous feedback - Each response can be added to a repository of verified responses. This way Seam can learn about user’s specific case and improve over time,
Explainability - Since the model is capable of independently analyzing the input and data, and draw conclusion based on its training we make sure that users can trace the entire decision-making process.
See more:
