I stumbled upon a situation when utilizing Coda AI’s table generation feature. It seems that there is an inconsistency between the intended column types and the actual ones generated, along with missing descriptions for specific columns. Here’s what I observed:
Steps to Reproduce:
Requested Coda AI to generate a table using the following specifications:
- Exercise Name: This column should be of type “Text” with a description that allows brief descriptions or instructions for each exercise.
- Muscle Group: This column should be of type “Selection List” with a description intended for specifying the targeted muscle group for each exercise.
Expected Result:
Ideally, the generated table should consist of two columns labeled “Exercise Name” and “Muscle Group,” each adhering to their respective column types and including the specified descriptions mentioned above.
Actual Result:
The AI-generated table does feature the “Exercise Name” column as intended. However, the “Muscle Group” column consistently defaults to a “Text” column, failing to match the designated “Selection List” type. Furthermore, the descriptions for both columns are inexplicably absent from the generated table.
Impact:
These inaccuracies regarding the column type for “Muscle Group” and the omission of descriptions adversely affect the functionality and practicality of the generated table.
Suggested Solution:
To address this matter, I propose that Coda AI be refined to incorporate the capability to accurately determine the designated column types (“Text”, “Selection List”, etc) and integrate the accompanying descriptions during the table generation process. By incorporating these specifications, users can ensure the utmost precision and efficacy of their tables.