Relax: AI Won’t Trigger an Apocalypse Since They Can’t Program

Why You Shouldn’t Fear the AI Revolution: A Candid Reflection on My Coding Experience

In recent days, I’ve immersed myself in an experiment involving two artificial intelligence platforms—ChatGPT and CLAUDE. What began as a straightforward task quickly transformed into an unexpected journey. My goal was to utilize these AI tools to create graphic files for my Substack newsletter. Initially, I considered using PowerPoint for this project, but I thought, “Why not explore the possibilities with AI coding?”

While that presentation would have typically taken a couple of hours in PowerPoint, I was eager to streamline the process through AI. Thus, with optimism, CLAUDE and I embarked on this ambitious endeavor to produce a polished set of graphics. The prospect of saving time filled me with excitement!

However, the experience soon evolved into a chaotic struggle. For every issue we managed to tackle, a whole new set of challenges emerged, leading to what felt like an endless game of whack-a-mole. Without diving into the overwhelming specifics, let’s just say that my favorite descriptor for the situation was “cluster—.” After numerous frustrating days, I ultimately decided to revert to PowerPoint and created a well-crafted presentation in just over three hours.

This experience reinforced a notion I hold about AI coding: right now, AIs function similarly to novice developers facing their first significant project. There’s potential for growth, particularly through methods like Reinforcement Learning from Human Feedback (RLHF) and RLHF-AI, where human experts help refine and enhance AI capabilities.

Nevertheless, the widespread anxiety surrounding artificial intelligence seems to be exaggerated or amplified by those with ulterior motives. Rather than succumbing to fear, I advocate a more relaxed approach. Pour yourself a glass of Scotch, take a couple of aspirin, and set your worries aside. After all, we’re likely far from an AI apocalypse, no matter how captivating the headlines may suggest.

Embracing the current limitations of AI can be a more productive route than letting fear dictate our perspectives. Let’s focus on how humans and machines can collaborate effectively, rather than worrying about an uncertain future.