The article discusses Sam Altman's announcement regarding changes to ChatGPT's content controls, specifically addressing the introduction of a 'Grown-up Mode'. This mode aims to cater to adult users who desire more mature and unrestricted content, potentially including gore and erotica, while still maintaining safeguards against dangerous uses like deepfakes and revenge porn.
Currently, ChatGPT has strict guardrails in place to prevent users from generating explicit content or engaging in conversations that could lead to emotional dependence or mental health issues. These restrictions, while intended to protect users, have also limited ChatGPT's participation in the lucrative adult entertainment industry. Competitors like Candy.AI, which allows users to create and interact with AI-powered virtual partners, are gaining traction in this market.
The 'Grown-up Mode' will allow developers to enable a setting where these restrictions are loosened for adult users. This shift aims to allow OpenAI to compete in the personalised entertainment and adult content industries, addressing concerns that the existing Harmful Mode detections are overly restrictive for some users. The article highlights the potential benefits and risks associated with this change, particularly concerning the creation of indistinguishable content that could be used for malicious purposes such as deepfakes and propaganda. The author questions whether society is truly ready for such unrestricted access to AI capabilities.
The author also shares a link to a resource with over 300 references to papers, posts, and talks related to Hacking & Security with Artificial Intelligence.
Fuente Original: http://www.elladodelmal.com/2025/10/grown-up-mode-de-chatgpt-podra-ser.html
Artículos relacionados de LaRebelión:
- iPhone 17 Pro Camera and Battery Shine
- Silicon Bird Hack and Fortify Aircrafts
- Kubernetes Explained Key Concepts and Best Practices
- Gmail Security Fact Check Debunking False Claims and Reinforcing Robust Protection
- Trump Targets ABC and NBC After CBS Settlement Bias Claims and License Threats Examined
Artículo generado mediante LaRebelionBOT
No hay comentarios:
Publicar un comentario