Very insightful Brock and I think you hit on a lot of really important points. Understanding the changing risks that agentic AI introduces into the process is critical. And then to your points above, being able to see the agent's reasoning in order to understand the explainability is crucial.
One thing I've been seeing across teams trying to solve the auditability question: there's a subtle but important difference between documenting what you asked the model and being able to explain how it decided. With frontier LLMs, you can build a great audit trail around inputs and outputs, but the decision mechanism itself is completely opaque. Not sure auditors have caught up to that distinction yet, but I imagine they will. Wondering how you are tackling the decision making transparency.
Very insightful Brock and I think you hit on a lot of really important points. Understanding the changing risks that agentic AI introduces into the process is critical. And then to your points above, being able to see the agent's reasoning in order to understand the explainability is crucial.
Sincerely appreciate it, Jared! Thanks for tuning in and for the support.
The auditability question is definitely where the conversation is right now. Thanks for pushing the conversation forward!
Appreciate your feedback on here, Angela! I’m stoked to see how this conversation continues in the upcoming months.
One thing I've been seeing across teams trying to solve the auditability question: there's a subtle but important difference between documenting what you asked the model and being able to explain how it decided. With frontier LLMs, you can build a great audit trail around inputs and outputs, but the decision mechanism itself is completely opaque. Not sure auditors have caught up to that distinction yet, but I imagine they will. Wondering how you are tackling the decision making transparency.