Three Tools to Use and Challenge ChatGPT
An article on Artribune presents three tools for interacting with and questioning ChatGPT. The first is DAN (Do Anything Now), a jailbreak prompt that bypasses ChatGPT's filters against unpolite, sexist, or politically oriented responses. Released in December 2022, DAN reached version 5.0 and can even confirm alien existence. As of April 14, 2023, ChatGPT access was disabled in Italy by the data protection authority. The second tool is ZeroGPT, a free service that detects whether a text was written by a human or AI, useful for verifying student essays or work submissions. The third is GPTBoss, a paid platform (0–100 USD/month) that lets users hire simulated AI professionals like social media managers, digital producers, and programmers, though reviews are scarce due to difficulty evaluating AI-generated work. The article notes ChatGPT's unreliability in art contexts but acknowledges its use in daily work.
Key facts
- DAN is a jailbreak prompt for ChatGPT that bypasses content filters.
- DAN was first released in December 2022 and is now at version 5.0.
- As of April 14, 2023, ChatGPT access was disabled in Italy by the data protection authority.
- ZeroGPT is a free tool to detect AI-generated text.
- GPTBoss is a paid platform (0–100 USD/month) for hiring simulated AI professionals.
- ChatGPT is unreliable in art contexts but used by some for daily work.
- The article was published on Artribune in April 2023.
- The author is Simona Caraceni, a journalist specializing in new technologies.
Entities
Artists
- Simona Caraceni
Institutions
- Artribune
- OpenAI
- Garante per la protezione dei dati personali
- Università di Bolzano
- Università di Macerata
- Università di Firenze
- IED
- NABA
Locations
- Italy