Member-only story
ChatGPT’s Enchanted Code: A Developer’s AI Quest
The adventures of coding with AI, and why human verification is more crucial than ever.
Once upon a time, in a coding world far, far away, Pietro Schirano, design lead at Brex, met the mystical, magical ChatGPT. This AI genie, a product of the GPT-4 architecture, astonished Pietro with its uncanny ability to generate code for classic games like Pong, Asteroids, and Breakout. Pietro was in awe, but he couldn’t shake the unsettling feeling that there was more to this AI than met the eye.
As ChatGPT continued to assist Pietro and his colleague, Ammaar Reshi, with coding projects, their amazement grew. The AI knew what code to generate, and it could even learn about games it had never encountered before. But as the duo worked with ChatGPT, they realized a crucial truth: no code, whether generated by human or machine, should ever be trusted without verification.
This tale, my dear readers, serves as a cautionary lesson for all those who dare to dabble in the realm of AI-generated code. Just as any fairytale has its twists and turns, so does the world of AI and coding.
The Dark Side of the AI Magic
Tony Smith, CTO at Rightly, warns that the risks of using ChatGPT for software development…
