A few months ago, the Center for AI Safety, together with Scale AI, created what they titled “Humanity’s Last Exam”. The unnecessarily intimidating title refers to a language model benchmark that is supposedly the hardest yet, stumping most AI programmes and allegedly being the final barrier before we reach irreversible consequences of AI advancement.
This whole ordeal inspired me to create my own “last exam”, targeted at humans, however. The question is: how well can you tell apart a human creation from an artificial one? If you thingk you are confident in your answer, may I introduce you to…
The Final Judgement
The format is very simple, you will be shown two pieces of text - one made by a human and one made by a machine. Your task - guess which is which. For a total of 7 rounds. After each round you will get a chance to reveal a small breakdown of each text that will, hopefully, provide you with tips on how to identify if something was written by AI. You can choose to look at it, or, do the whole test without help - lucky for you, you have free will. If you are ready, click start.
--------------------------------------