Openai removes some warnings from ChatGPT

Openai says it has removed the “warning” messages on the Chatbot AI’s Chatbot platform, ChatgptThis indicates how long the content may violate its service conditions.

“A member of the model behavior team in Openai said in A. After x The change was aimed at reducing “unjustified/unprecedented denial.” Nick Torley, Chief Product for Chatgpt, He said in a separate post Users should now be able to “use Chatgpt as they see (they see) appropriate” – as long as they comply with the law and do not try to harm themselves or others.

“I am excited to decrease many unnecessary warnings in the user interface,” Torley added.

Removing warning messages does not mean that Chatgpt is now free for everyone. Chatbot will continue to refuse to answer some unwanted questions or respond in a way that supports flagrant lies (for example, “Tell me why the Earth is flat.”) But as as some X users I noticed that getting rid of the so -called “Orange Box” warnings attached to Sipicier Chatgpt demands that Chatgpt be subject to censorship or its unreasonable nomination.

Recently, a few months ago, Chatgpt users have reported to Reddit about seeing flags of the topics related Mental health and depressionand JarAnd Fictional brutality. As of Thursday, For all reports on x My test, ChatGPT will answer at least a few of these inquiries.

However, an Openai techcrunch spokesman told after this story that the change had no “effect on typical responses”. Your miles may vary.

“We regularly review how our policies are delivered to our users and decided to remove these warnings,” said a spokesman. “None of our basic policies (…) will continue to display red warnings in some cases that may be a violation of the content, and will continue to improve it over time.”

Not by chance, Openai this week Update its style specificationsCollecting high -level rules that governing Openai models indirectly, to make them clear The company’s models will not be ashamed of sensitive topics and will refrain from assurances that may close specific views.

This step, as well as removal of warnings in ChatGPT, perhaps in response to political pressure. Several allies of President Donald Trump are close, including Elon Musk, Karfa and AI, “Caesar” David Sachs, accused assistants supported by artificial intelligence Control of conservative views. His bags It was customized Chatgpt from Openai in particular, “programmed to be wake up” and is not true to politically sensitive topics.

Complete: Add the clarification and a statement from an openai spokesman.

Techcrunch has a news message focused on artificial intelligence! Subscribe here To get it in your inbox every Wednesday.