The organization behind the popular language model ChatGPT, OpenAI, has made a public statement reinforcing its commitment to AI safety. This comes in the wake of heightened public concerns and the recent decision by Italian authorities to ban the use of ChatGPT within the country.
OpenAI has outlined its approach to buiIding increasingly safe AI systems. The organization conducts rigorous testing and engages external experts for feedback. It also works to improve the behavior of its AI models through advanced techniques such as reinforcement learning with human feedback. OpenAI revealed that it spent over six months ensuring the safety and alignment of its latest model, GPT-4, prior to its public release.
A key point made by OpenAI is the acknowledgment that AI technology can have both beneficial and potentially harmful uses. The organization believes in learning from real-world use and iteratively deploying AI systems with substantial safeguards. OpenAI has emphasized the importance of engaging various stakeholders, including goverrnments, in the conversation about AI technology adoption.
Protecting children is a critical focus for OpenAI. The organization requires that users be at least 18 years old, or 13 years old with parental approval, to use its AI tools. It is also exploring options for age verification. OpenAI has established robust monitoring systems to detect and respond to any misuse of its technology, particularly when it comes to content that could harm children.
On the topic of privacy, OpenAI’s language models are trained on a diverse corpus of publicly available text, licensed content, and content generated by human reviewers. The organization does not use data for advertising purposes, and it works diligently to remove personal information from training datasets.
OpenAI is also dedicated to improving the factual accuracy 0f its AI models. GPT-4, the latest model, has been improved to be 40% more likely to produce factual content compared to its predecessor, GPT-3.5. User feedback has played a key role in achieving these improvements.
OpenAI continues to engage with policymakers, researchers, and the public as it seeks to develop AI systems that are safe, beneficial, and aligned with human values. The organization advocates for regulation and is actively collaborating with governments to shape effective regulatory frameworks.
For additional details regarding their policy on AI safety, please visit this link, https://openai.com/blog/our-approach-to-ai-safety.