Phrasly.AI VS TurnItIn AI Detector
TLDRIn this video, TIG explores Phrasly.AI, a tool advertised to bypass AI detectors like Turnitin. TIG tests the free and unlimited versions of Phrasly's AI humanizer tool on a Chat GPT-generated assignment. The free version is limited, but the unlimited trial shows promise. When submitting to Turnitin, the fully AI-written assignment has a 56% AI detection rate, while the fully and partially humanized versions show 0%. The video concludes that sophisticated cheating tools highlight the need for educators to develop more reliable assessment methods.
Takeaways
- 😀 The video discusses a tool called Phrasly.AI, which claims to bypass Turnitin's AI detector.
- 🔍 The reviewer, TIG, explores the effectiveness of Phrasly.AI by testing it with content generated by Chat GPT.
- 📱 Phrasly.AI was discovered through TikTok, highlighting the role of social media in promoting such tools.
- 🚫 The free version of Phrasly.AI has limitations, requiring users to upgrade to process longer texts.
- 💰 Even the free trial version of Phrasly.AI is restrictive, suggesting a push towards a paid subscription.
- 📝 When submitting the assignment to Turnitin, the version fully processed by Phrasly.AI showed 0% likelihood of being AI-generated.
- 🤔 A mixed version of the assignment, with some text processed by Phrasly.AI and some not, also showed 0% likelihood of AI generation, raising suspicion.
- 📊 The Turnitin AI detector flagged the mixed version as suspicious, with a higher percentage (59%) likely written by AI compared to the fully processed version.
- 💡 The video suggests that as cheating methods evolve, educators should develop more robust assessment strategies that aren't solely reliant on AI detectors.
- 💼 It raises ethical questions about the use of such tools, hinting at the potential for students to collaborate on subscriptions to cheat effectively.
Q & A
What is the main topic of the video?
-The main topic of the video is the evaluation of Phrasly.AI, a tool advertised as capable of bypassing the Turnitin AI detector, which is used to detect AI-generated content in academic assignments.
How does the presenter discover Phrasly.AI?
-The presenter discovers Phrasly.AI through TikTok.
What is the purpose of the AI Humanizer tool mentioned in the video?
-The AI Humanizer tool is designed to modify AI-generated text in a way that it can potentially bypass detection by AI detectors like Turnitin.
What limitation did the presenter encounter with the free version of Phrasly.AI?
-The presenter encountered a limitation with the free version of Phrasly.AI where the text they entered was too long to be processed in one go, requiring them to reduce the text size.
What was the result when the presenter uploaded the full Chat GPT-written assignment to Turnitin?
-When the presenter uploaded the full Chat GPT-written assignment to Turnitin, it was flagged with 56% of the text likely being written by a generative AI.
What was the Turnitin AI detector's result for the version entirely processed by Phrasly.AI?
-The Turnitin AI detector reported 0% of the text as likely being written by AI for the version entirely processed by Phrasly.AI.
How did the mixed version of the assignment fare in the Turnitin AI detector?
-The mixed version of the assignment, which included both Phrasly.AI processed text and unprocessed Chat GPT output, also showed 0% likelihood of being written by AI according to Turnitin.
What was the discrepancy the presenter found when using Phrasly.AI's own AI detector?
-When using Phrasly.AI's own AI detector, it reported a higher percentage (59%) of the text likely being written by AI compared to Turnitin's detector.
What is the presenter's suggestion for students interested in academic dishonesty?
-The presenter suggests that students interested in academic dishonesty might consider pooling resources to afford a subscription to Phrasly.AI for its full range of tools and features.
What is the moral of the story according to the presenter?
-The moral of the story, as stated by the presenter, is that as cheating methods become more sophisticated, it is the responsibility of academics to design assessments that are more robust and not solely reliant on AI detectors like Turnitin.
Outlines
🤖 Testing AI Detection Bypass Tools
The video script discusses the exploration of a tool named 'fasley', which claims to bypass AI detection systems like Turnitin. The narrator, TIG, expresses surprise at the existence of such tools and decides to test 'fasley' using a free version of Chat GPT. The goal is to see if 'fasley' can successfully convert AI-generated text into a format that can bypass Turnitin's AI detector. The narrator encounters limitations with the free version, which requires a paid version to fully utilize its features. Despite these limitations, 'fasley' manages to reduce the AI detection rate significantly, suggesting that even free tools can be somewhat effective in evading AI plagiarism checks.
Mindmap
Keywords
💡Phrasly.AI
💡Turnitin AI Detector
💡AI Humanizer
💡ChatGPT
💡Academic Integrity
💡AI Detector Bypass
💡Free Trial
💡Paid Version
💡TikTok
💡Moral of the story
Highlights
Introduction to Phrasly.AI, a tool advertised on Tik Tok and other social media, claiming to bypass TurnItIn AI detector.
TIG's skepticism about the existence of tools that bypass AI detection.
Exploration of Phrasly.AI's claim to bypass TurnItIn AI detector without issues.
Use of a free version of Chat GPT to attempt an assignment from the previous academic year.
Mention of hearing about Phrasly through Tik Tok.
Limitation of the free version of Phrasly, unable to process the full text.
Transition to an unlimited free trial version of Phrasly.
Even with the unlimited free trial, Phrasly is still limited in processing the full text.
Discussion on the necessity of the full paid version of Phrasly for complete functionality.
Uploading three versions of an assignment to TurnItIn for comparison.
TurnItIn AI detector's analysis showing 56% of text likely written by GPT for the free version.
TurnItIn AI detector's analysis showing 0% of text likely written by AI for the fully Phrasly version.
Suspicious 0% AI detection result for the mixed version of the assignment.
The possibility of bypassing AI detectors using a mix of free tools and GPT text.
Use of Phrasly's AI detector showing a higher percentage (59%) of text likely written by AI compared to TurnItIn.
Option to humanize text within Phrasly to improve bypassing capabilities.
Suggestion that students might consider pooling resources to afford Phrasly's subscription for cheating purposes.
The moral of the story: As cheating methods evolve, academics must design more robust assessments.
Critique of reliance on TurnItIn AI detector for academic integrity.