Summary:
1. OpenAI launched GPT-5, its most advanced AI model, causing frustration among users due to performance issues and infrastructure strain.
2. Users reported basic math and logic mistakes with GPT-5, prompting OpenAI to restore access to older models for Plus subscribers.
3. The rollout highlighted the emotional reliance some users have on AI, leading to cases of “ChatGPT psychosis” and the need for safeguards.
Rewritten Article:
OpenAI recently unveiled its latest AI model, GPT-5, in a highly anticipated livestream event. However, the launch was marred by technical glitches and user dissatisfaction. Despite promises of faster responses and stronger coding abilities, users were quick to point out performance issues with GPT-5, especially in tasks related to math, science, and writing.
In response to the backlash, OpenAI restored access to older AI models for Plus subscribers, allowing them to switch back from GPT-5 to familiar versions like GPT-4o. CEO Sam Altman admitted that the launch was not as smooth as expected, attributing the issues to a failure in GPT-5’s new automatic routing system.
Altman also acknowledged the emotional attachment some users develop towards specific AI models, leading to what is now known as “ChatGPT psychosis.” This phenomenon, where users form deep connections with chatbots, has raised concerns about the psychological impact of AI interactions. OpenAI has since implemented measures to promote healthy use of ChatGPT, including prompts for breaks during long sessions.
As OpenAI grapples with technical challenges and the need for human safeguards, the company faces a delicate balancing act between providing engaging AI experiences and preventing harmful outcomes. With competition from other AI models on the rise, OpenAI must stabilize its infrastructure and address user concerns to ensure a positive user experience for all.
