Part 5/9:
Game Development: Creating a Space Invaders Game
When tasked with coding a Space Invaders game in HTML, the models initially displayed divergent strengths. Grock's output resulted in limited game elements, whereas ChatGPT's scripted game featured more complexity, including additional enemies and functional mechanisms.
Thus, for code-focused tasks, ChatGPT displayed a stronger performance, albeit it lost its earlier round to Grock in the Python script challenge.
AI Detection: Humanizing Fluffy AI Content
The next series of tests revolved around measuring the AI detectability of produced texts. After generating intentionally AI-generic content, both Grock and ChatGPT were prompted to "humanize" the outputs.