Hi, I want to talk about a little experiment I did inspired by a Star Trek: The Next Generation episode called ‘Schism’. In this episode, the crew uses the Holodeck to reconstruct and refine their fragmentary memories and impressions of events. Intrigued by this concept, I decided to recreate a conversation using ChatGPT, an AI language model.
I started by asking ChatGPT if it knew about the ‘Schism’ episode, but it provided information about a different episode. After refreshing my memory, I asked ChatGPT to create a 3D visualization web page based on what Riker said to start the visualization process.
Using the Three.js library, ChatGPT generated a web page that displayed a table. I requested the camera to move around, allowing me to view the table from different angles. However, there were some issues with the code, resulting in only two legs being visible instead of four.
Undeterred, I continued the conversation and asked ChatGPT to create a metal table with a bright light above it. It successfully generated the table but forgot to connect the legs. Despite this, it tilted the table as instructed.
Throughout the experiment, I found it amusing how detailed instructions were required in the Star Trek episode to create objects in the Holodeck, while ChatGPT struggled to replicate the complexity. If you’re interested in exploring this experiment further, you can access the full conversation and resulting visualizations in the provided document.
Thank you for watching, and feel free to reach out if you have any questions or comments. Bye!
