From Words to Wonders: The Magic of Prompt Engineering Unleashed

I always thought that Arthur C. Clarke’s third law belongs to the realm of science fiction: “Any sufficiently advanced technology is indistinguishable from magic.” However, I have to question this assumption due to recent developments in AI and prompt engineering. First, Prompt engineering was a science. It had rules, best practices, and formulas that worked or worked most of the time. Then, if you practiced enough and were creative enough, it could become an art. The more intuition, imagination, and innovation you had, the better results you could accomplish. 

But recently, prompt engineering ventured from science and art to the magical territory, at least to my mind and most humans. The verified fact that framing a prompt as a quest in the fictional world of Star Trek increased the quality of the results is hard to explain or distinguish from magic. The title of the scientific paper already covers it well: “The Unreasonable Effectiveness of Eccentric Automatic Prompts.” The researchers prompted the AI to create its prompt and improved its creation by wording it as a script for the TV above show.

Studies show how tipping the AI, being polite, or providing emotional stimuli can enhance its answers. It is not hard to see that we are approaching a time when the anthropomorphization of AI will be so prevalent that it cannot be stopped. We can have hypothetical, philosophical discussions about whether they are conscious or have passed the Turing test. But the real test is not of the machines but of the humans. It is easy to see that a significant percentage of the population will treat AI as a de facto conscious entity. By their nature, it is a different kind of intelligence, which, although developed by humans, is already hard to distinguish from magic. So it is not an alien technology that Clarke’s third law will apply to, but humans’. And very soon.