AI-based large language models (LLM) are trained on vast quantities of public information to connect the dots and output knowledge about various Global Problems.
However, regarding local problems, the current AI/LLM tools, lacking local training data, are ineffective. In such cases, we still need to rely on groups’ collective knowledge and creativity.
The table categorizes complex problems into global and local types, indicating their relevance to AI/LLM.
Global problems are AI/LLM relevant with significant AI/LLM content, require minimal manual intervention, and are helpful for IdeaGens. Local problems are not AI/LLM relevant, have no AI/LLM content, and require full manual intervention, but are still helpful for IdeaGens.
Creating local LLM is a costly and time-consuming task. Therefore, one will continue to require that Stakeholders are continuously engaged and queried to build a complete local knowledge base.
