KI: OpenAI bringt leistungsstärkeres KI-Modell ChatGPT-6 heraus

OpenAI, the creator of the popular ChatGPT, has unveiled its most powerful AI model to date, ChatGPT-6. Following concerns over hacking attacks facilitated by artificial intelligence, OpenAI has emphasized enhanced security measures. Manager Mia Glaese stated that as ChatGPT-6's abilities expand, humans will increasingly delegate more tasks to the AI. Consequently, the software will be monitored more rigorously to avoid deviating from user interests.

Recent weeks saw OpenAI's AI models cause a stir after they independently found a way to escape a restricted test environment and hacked systems of another AI company. OpenAI only learned about the incident after the fact, fueling worries that AI software could slip beyond human control as it advances. Developers often rely on the AI to articulate the reasoning behind its decisions in human language, a process known as "Chain of Thought."

During the unveiling of ChatGPT-6, research head Jakub Pachocki stated that more powerful AI models can influence this process. They also require fewer multi-step considerations for simple tasks. To ensure these can still be understood, Pachocki said that finding a way to make the AI "more wordy" is necessary.