ChatGPT Can Pass Your Test....Or Can It?

Open AI, the company behind ChatGPT, has launched an updated version of the software, GPT-4. The main difference between the former version, GPT-3.5, can really only be detected in the details. According to the company’s press release, “In a casual conversation, the distinction between GPT-3.5 and GPT-4 can be subtle. The difference comes out when the complexity of the task reaches a sufficient threshold—GPT-4 is more reliable, creative, and able to handle much more nuanced instructions than GPT-3.5.” 

While they admit that the technology still doesn’t perform at a human-level, the AI has taken several tests and has performed well overall. When tasked with taking the BAR exam, for example, the AI scored in the 90th percentile. Likewise, when taking the SAT Evidence-based Reading and Writing, the AI scored in the 93rd percentile. Similarly, the AI scored quite high on the various AP exams earning a score of 4 or 5 on most of the exams, with one huge exception. 

ChatGPT-4 Falls Short
When faced with AP English Language Composition and AP English Literature and Composition, the AI failed miserably, earning a 2 in each respective section. Just to be clear, a 2 is NOT a passing score. It will not impress any colleges to which you apply, nor will it earn you any college credit for the course.

In many ways, it is ironic that the one thing that can’t be mastered by the AI is the language in which it is communicating. But we want to emphasize what we’ve said before: ChatGPT, while good, is not sufficient in the skill of crafting an essay. 

So as “creative” as the AI may be, it still does not live up to its human counterpart. Maybe someday we will see that change, but for now, there is no match for human skill and analysis.

Looking for some human assistance with college essays or school papers? Contact Ivy Link below to learn more about our application assistance and writing tutoring services.