CoPilot - ChatGPT & the Tekla Structures API - Super Human Detailer
This video shows how Language Learning Models (LLMs) such as CoPilot and ChatGPT have really reached a new level. As an expert programmer of the Tekla Structures .NET based API using the C# programming language, I was amazed how after 3 iterations of prompts and small guided adjustment recommendations in human readable English yet still in Tekla steel detailer lingo, I was able to get the resultant code of my prompts. If I can create code using prompts, I can create job or task specific coding from a verbal narrative that can essentially automate modeling or drawing API tasks to do work for me. In the near future with a network of AI Agents and training from my code and prompt data, the term "Super Human Detailer" could start to become a reality no matter where you are located in the globe and what size your detailing firm is. The more experience and data you have in your back pocket makes these tools even more valuable.
Imagine the long term future. Pass AI agents and LLM's Tekla Structures models and drawings with a headless API in the cloud who knows where drawing cleanup, connection application and design, and model integrity checking can go.
From there, looking at scenarios such as trends on efficient usage or common stocks of main and connection materials used on projects. Maybe identify inefficiencies or trends in design and fabrication for purchasing, estimating, and routing in the shop. Interesting...
Chapters
00:00 - Introduction and My Background
02:03 - Prompt 1 Two Columns and a Beam
04:24 - Prompt 2 Connect a Column and Beam with 141
05:44 - Iteration 1 Use the Connection Class
06:35 - Iteration 2 Change Properties on the Connection Class
08:31 - Iteration 3 Change What is Passed to the LoadAttributes Method