• Cover
  • News
  • Entertainment
  • Writing + Books
  • Sports
  • School
  • Science and Technology
Menu

Park Forest Times

  • Cover
  • News
  • Entertainment
  • Writing + Books
  • Sports
  • School
  • Science and Technology

Image Source: Unsplash

The Strawberry That Tricked A.I. by Tara Gopalan

September 16, 2024

AI is one of humanity’s most complex (and scary) creations. It can simulate human speech, retrieve information from all over the world if prompted, and even take on the role of specific people in history. All of this is great, but AI still isn't advanced enough to replace humanity. How do I know this, you ask? Just ask it a simple question: How many ‘R’s are in the word ‘strawberry’. You may find that AI isn't as smart as it seems.

Many of us have used ChatGPT, one of the most advanced search engines ever created. It’s reliable, right? Wrong. When asked how many ‘R’s there are in ‘strawberry’, it replies, “There are two R’s in the word ‘strawberry.’” When you try to count them yourself, you’ll find that there are 3 ‘R’s instead of 2. You may have thought that AI learns from the responses people give it, but when told that it counted wrong, the software becomes defensive: "You split up the word to make it seem like there are three R's, but when counting the actual letters in the word 'strawberry,' there are still just two R's (Eaton).” It refuses to admit it’s wrong! Now why is such an intelligent software so clueless when it comes to spelling?

Well, Chat GPT is programmed using Large Language Model algorithms, meaning that it’s fed large amounts of information at one time. This just means that it’s seen many different types of data, and can therefore identify what singular words mean. However, like in the case of “Strawberry”, it only knows the meaning of the word and can’t actually comprehend what it’s saying. Because of this flaw, ChatGPT spews out a lot of misinformation unbeknownst to its users! All of the information that ChatGPT stores is just a clump of code hidden in the software, and that’s hopefully all it will be for a long time. Thankfully, AI isn't yet advanced enough to have comprehensive thought, and humans are still needed in the data processing world. 

To conclude, AI isn't as reliable as we think it is. Unlike humans, it can’t learn and grow when fed new information. So before you use Chat GPT or any other AI to write your English paper or do your math homework, make sure to double check whether you have a strawberry or a stawberry. 

Work Cited

Eaton, Kit. “How Many R's in 'Strawberry'? This AI Doesn't Know.” Inc.com, Inc.com, 28 August 2024, https://www.inc.com/kit-eaton/how-many-rs-in-strawberry-this-ai-cant-tell-you.html. Accessed 6 Sep 2024.

← What to Expect for Pennsylvania in Winter 2024-25 by Liam ScheelerChatGPT: a helpful tool or a harmful crutch for learning? by Evelyn Patch →

Sports News

Sports
Huge Win for the Thunder as Shai Drops 40 By Miles Laskowski
Huge Win for the Thunder as Shai Drops 40 By Miles Laskowski
about 3 weeks ago
College Football Conference Ideal Realignment by Grant Potoczny
College Football Conference Ideal Realignment by Grant Potoczny
about 3 weeks ago
NBA Playoff Update #3 by Marshall Laskowski
NBA Playoff Update #3 by Marshall Laskowski
about 3 weeks ago
Wake Forest early exit in ACC Baseball Tournament  By: Andrew Zajac
Wake Forest early exit in ACC Baseball Tournament By: Andrew Zajac
about 4 weeks ago
NBA Playoffs Update Number Two by Marshall Laskowski
NBA Playoffs Update Number Two by Marshall Laskowski
about a month ago

Calling All Creative Writers!

What to share some creative writing or become a published writer with an audience? Simply share your work with Mr. Cunningham, Mr. Harper, or Mr. Cimino and we'll work on publishing your material right here in the Park Forest Times.


Powered by Squarespace