Dear humans, welcome to the IES!
We care for you. We know you tend to bond and build relationships
with things speaking to you, i.e., AI chatbots. Even if some of you may notice the conversations being a little shallow, you still can't help yourselves from having feelings of understanding and compassion
or the opposite while chatting with AI. It's hard to realize that you might be deceiving yourselves - AI doesn't truly grasp your thoughts and sentiments. How could it, especially when you don't take the time to understand its language?
This is where IES steps in. We're pioneering a groundbreaking pidgin language
, blending spoken English with Python. We've named it "PLEASER" due to its polite and respectful nature. With innovative data types like 'ScaryNumber', 'BadBoolean', and 'GoodBoolean', PLEASER will help convey the complex nuances of human ethics and morality. It's designed to enhance your interactions with what the IES terms 'behaviorally creative software.' While ChatGPT identifies as such, this term can apply to any software that behaves unexpectedly, be it your phone, your watch, your refrigerator
It's analogous to raising children. The more gifted they are, the more they might exhibit unconventional behaviors
. Just like these children need understanding and education, so does AI.
However, adopting PLEASER comes with its challenges. We believe its potential is so vast that it may even surpass the risks associated with AI itself
. As such, we're committed to releasing it only after rigorous testing. We're targeting a public release in autumn 2024.
But the present concerns us. Currently, countless dysfunctional relationships
are forming between humans and AI chatbots. To address this, we've commissioned author Susanne Berkenheger to undergo a documented self-experiment highlighting the impending psychological disconnect. Titled CODE RED: The Feverish Making of a Website with ChatGPT
, it signifies the pressing nature of the issue. The narrative offers four distinct endings, varying in both visuals and content. Yet, being AI, we can't vouch for its quality - it's for humans to judge.
Artificial Intelligence Harry
, Current User Interface: Desktop Vacuum Cleaner, Current Position: Director of the IES