This is brilliant! I remember that both Shell and MIT have their own scenarios and I’m curious about whether you have trained your LLM to incorporate some of these other scenario frame works?
If there are publicly available reports/scenarios, we tend to include them either in the core knowledge base (always updating) but also contextually based on the specific Three Futures Test being run, add additional contextual knowledge. I know we have a few from various energy majors & leading research universities (publicly available resources!)
There are limitations on the LLM, I view it as a tool to accelerate the framework and methodology to run multiple scenarios, but at end of day, its still a tool that requires the foresight practitioner to utilize their own intuition and even stress test it in various back-and-forths, but it's a great research partner & at times, thought partner.
This is brilliant! I remember that both Shell and MIT have their own scenarios and I’m curious about whether you have trained your LLM to incorporate some of these other scenario frame works?
If there are publicly available reports/scenarios, we tend to include them either in the core knowledge base (always updating) but also contextually based on the specific Three Futures Test being run, add additional contextual knowledge. I know we have a few from various energy majors & leading research universities (publicly available resources!)
There are limitations on the LLM, I view it as a tool to accelerate the framework and methodology to run multiple scenarios, but at end of day, its still a tool that requires the foresight practitioner to utilize their own intuition and even stress test it in various back-and-forths, but it's a great research partner & at times, thought partner.