Recently, over 30 companies released their own AI content detector models. But with this many tools it can be challenging to know which detectors to trust. Currently, most of these detectors are simple replicas of the GPT2 Output Detector.
So in this post I will summarize the top 5 unique AI detection tools. These are the best software tools to help you identify whether some content appears AI-generated.
1. GPT2 Output Detector
This is the father of most current detectors. It’s a 2019 supervised ML model trained by OpenAI to classify whether some text is GPT2 generated. The tool is open source and it’s trained to detect GPT2 content on 1.5GB of data.
This detector is straightforward to use. All you need to do is take your text and plug it into the detector. The detector will respond with a probability that the content is real or fake.
The tool provides you with a probability that your text is fake. However, it doesn’t give you a descriptive measure of what you can fix. That’s why it’s best to paste one paragraph at a time to this detector to see what sections you need to change.
WARNING: Fake Duplicates
Companies are taking this GPT2 Output Detector and marketing it as their own unique ChatGPT/GPT3 detector. There is no advantage to using those…
