Model Hubris: On the Presumptuousness of Large Language Models

We can expect large language models to be deployed in numerous contexts where a friendly, natural-language interface is expected. In order to augment their functionality, we might expect these systems to interface with other AI or software agents, databases, and systems.

In these cases it would be expected that the large language model would:

• Apply common sense reasoning
• Be able to perform simple logical inferences
• Not make assumptions that are inappropriate to the task

This last point - presumption - is the focus of this report.