> LLMs are probabilistic and non-deterministic

This is a polite way of saying unreliable and untrustworthy.

The problem facing enterprise is best understood by viewing LLMs as any other unreliable program.

> We’ve found that treating LLMs as suggestion engines rather than decision makers changes the architecture completely.

Sure does. Look at the damage LLM "suggestions" are doing to open source projects wordwide.