Want to sneak your AI generated text in under the radar? Try UndetectableGPT, an AI Humanizer that’s available for free online. That’s kinda funny, right? An AI making something sound more human.
So I decided to give UndetectableGPT a try with a real response. First, let me give you some background
In response to FunnyMonkey’s post about AI in Education, I searched Perplexity to see how it might respond. Before I show you Perplexity’s modified response
FunnyMonkey’s Critique
FunnyMonkey was responding to an article about bias in large language models. Here’s an excerpt from the paper’s abstract and it IS an interesting read:
…language models embody covert racism in the form of dialect prejudice, exhibiting raciolinguistic stereotypes about speakers of African American English (AAE) that are more negative than any human stereotypes about African Americans ever experimentally recorded.
By contrast, the language models’ overt stereotypes about African Americans are more positive. Dialect prejudice has the potential for harmful consequences: language models are more likely to suggest that speakers of AAE be assigned less-prestigious jobs, be convicted of crimes and be sentenced to death.
Finally, we show that current practices of alleviating racial bias in language models, such as human preference alignment, exacerbate the discrepancy between covert and overt stereotypes, by superficially obscuring the racism that language models maintain on a deeper level.
In sharing the link, Bill Fitzgerald (FunnyMonkey) shares these accompanying comments:
If you are advocating for AI in education, you are actively advocating for multiple forms of harm. You might have good intentions. You might not mean to. And that is completely irrelevant.
Impact matters more than intent, especially in situations like schools where existing power imbalances are already at play.
So, what did Perplexity AI say?
Perplexity.ai’s Response
I asked Perplexity to respond, then told it to cut the response down. This got a page long response down to 3-4 sentences:
Introducing AI in education can have unintended consequences, like reinforcing existing inequalities. To avoid harm, it’s crucial to design inclusive tools, detect biases, and support teachers. By prioritizing student well-being and equity, AI can be a valuable addition to education.
I can see someone writing that text, but I’m curious to see UndetectableGPT’s approach.
UndetectableGPT
Here’s what UndetectableGPT came up with…I’m not all that impressed:
Introducing AI into education comes with the risk of unintentionally deepening existing inequalities. To minimize these harms, it’s essential to create inclusive technologies, actively detect and correct biases, and provide robust support for educators. By keeping student well-being and equity at the forefront, AI has the potential to significantly enhance the educational experience.
So, what can you do?
ChatGPT Version
I decided to ask ChatGPT 4o to write this like a third grader…see results below:
Prompt
Write a very simplified and short explanation of the following text. Use only short sentences without commas or conjunctions, and keep the language at a 3rd-grade reading level:
‘Introducing AI in education can have unintended consequences, like reinforcing existing inequalities. To avoid harm, it’s crucial to design inclusive tools, detect biases, and support teachers. By prioritizing student well-being and equity, AI can be a valuable addition to education.
Result
AI in schools can cause problems. It can be unfair. We need fair tools. We must check for mistakes. Teachers need help with AI. This way, AI can help everyone learn.
Other AI Humanizing Tools
Here are some AI Humanizing tools that I found worthwhile, if not as good as a full-blown chatbot like ChatGPT: