Tag: is retrieval-augmented generation

Using Knowledge Elicitation Techniques To Infuse Deep Expertise And Best Practices Into Generative AI

The post Using Knowledge Elicitaticom. Being innovative and using knowledge elicitation techniques to infuse best practices and deep domain expertise into generative AI and LLMs. getty In today’s column, I address how you can discover hidden best practices that underlie deep expertise and then codify that secret sauce of domain knowledge into modern-day generative AI and large language models (LLMs). The crux of this vexing matter is that we can lean into the precepts of knowledge elicitation that were well-formulated during the rules-based expert systems era. I realize that some who are steeped in LLMs might balk at using the “dated” approaches from an earlier era of AI. Despite whatever adverse opinion someone has about so-called GOFAI (good old-fashioned AI), there is absolutely crucial value in leveraging tried-and-true techniques of eliciting deep domain knowledge. It can aid in turning generative AI and LLM into a bastion of best practices and deep expertise for a chosen domain. Let’s talk about it. This analysis of AI breakthroughs is part of my ongoing Forbes column coverage on the latest in AI, including identifying and explaining various impactful AI complexities (see the link here). Getting An LLM To Be An Expert Suppose that you want to turn an LLM into an expert within a particular domain. Perhaps you want to devise an LLM that has deep medical expertise and is especially proficient in urology or maybe neurology. Another possibility might be to shape generative AI to be on par with a fully qualified lawyer in real estate law or be a mental health therapist that is deeply versed in CBT (cognitive behavioral therapy). What can you do to shift a general-purpose LLM into being steeped in a specific domain? The usual approach consists of gathering as many documents as you can find that encompass the domain of interest. You.