Reality Check: Overcoming the Hallucination Hurdle in Generative AI

In September 2023 I spoke at the LLMs and the Generative AI Revolution virtual conference. You can find the talk below:

I discuss the topic of building reliable applications in domains where correctness matters on top of LLMs. I start by talking through techniques for working with the LLMs themselves, and also spend some time on talking about how to architect applications around the unreliability LLMs to ensure the software continues to maintain quality over time and can be improved with more usage.