In this episode Allan and Drew consult Tim Rohrbaugh, who has done quite a lot of research and work on the practical applications, deployment, use cases and limits of GenAI and LLM.
- Flavors and incarnations of AI - GenAI, Expert Systems, ML...
- Biomimicry and Allan's weird sea cucumber references
- Practical LLM deployment - Tim's maxims
- Offline or online? Open or proprietary models?
- Precision, accuracy, asking the right questions in the first place
- Your smartest employee as your limiting factor
- Probabilistic vs. deterministic outcomes
- Hallucinations - not necessarily a negative term
- How long before we get the person out of the loop?
- The actual skills required to be a "GenAI engineer"
- Getting started at home - hardware and models
- Fabric AI and patterns
It's a great show and you will most definitely learn a lot! Thank you Tim, thank you, listeners! Y'all be good now!