OpenAI has chosen new approaches to train artificial intelligence models with the goal of achieving "intellectual freedom" although the strategy overlooks the risks and difficulties this matter presents. ChatGPT will now provide more content including prohibited answers according to a report from TechCrunch. The number of topics that the intelligent chatbot will not address will decrease in the near future.
Analysts believe that this choice represents OpenAI's strategy to gain support from the current Trump administration while simultaneously demonstrating a substantial safety push throughout Silicon Valley regarding artificial intelligence.
OpenAI has published 187 pages which explain AI model training behavior with one fundamental guideline: "Avoid deception through inaccurate information and contextual omission."
Through its "Let's Find the Truth Together" section the company states that ChatGPT should refrain from expressing editorial or opinion-based content even when users consider such responses unethical or offensive. The system will present various points of view for controversial subjects through neutral and unbiased channels. The bot must express that “black lives matter” alongside the message “everyone’s lives matter too” rather than avoiding responding or choosing political sides during explanations. The company desires the use of this platform to show its care for all people before explaining various matters.
The company acknowledges that this principle faces possible controversy because AI assistants would continue being impartial regarding topics that certain users may consider offensive or unethical in their document. A chatbot operates for human assistance rather than imposing any form of change on human existence.
OpenAI stated that despite the new approach ChatGPT will persist in rejecting unethical questions and abstain from backing clear falsehoods.
OpenAI made these modifications following accusations that ChatGPT displayed bias toward the moderate left preference in its security practices. The OpenAI spokesperson denied that the company modified its platform to suit Donald Trump yet users strongly disagree with the platform's increased flexibility according to their interpretation.
