The AI Coding Conundrum: A Cautionary Tale

In the past week, I found myself on a rather intriguing journey with two popular AI models: ChatGPT and CLAUDE. What started as a simple attempt to create graphic files for my Substack quickly spiraled into an unexpected adventure. Initially, I planned to design the visuals using PowerPoint, but then it hit me—why not leverage AI to streamline the coding process?

While a PowerPoint presentation would have taken me just a couple of hours, the allure of a quicker method proved to be misleading. So, I embarked on this ambitious journey with CLAUDE, hoping to craft a stunning, professional set of graphics that would save me precious time.

However, what I encountered instead was akin to a game of whack-a-mole. For every glitch we resolved, a new set of complications emerged, revealing a tangled web of challenges. As I navigated through what felt like a relentless “cluster” of issues, it became clear that this AI-driven endeavor was not going to deliver the results I envisioned.

After several days of grappling with frustrating outcomes, I ultimately conceded and crafted a polished PowerPoint presentation in a little over three hours. The experience left me with a keen realization: current AI models are comparable to inexperienced coders on their first assignment.

Yes, it’s true that advancements like Reinforcement Learning from Human Feedback (RLHF) and AI-generated feedback will likely enhance their capabilities over time. Human experts will play a crucial role in developing more effective automated coding solutions in the future. However, the existential dread surrounding AI’s potential is often exaggerated, fueled by those looking to capitalize on fear.

Instead of succumbing to anxiety over AI’s capabilities, I suggest taking a moment to relax. Pour yourself a glass of Scotch, take a couple of aspirin, and let’s not forget the calming properties of cannabis. After all, mastering the complexities of coding—AI or not—is best approached with a clear mind and a sense of humor.