Servers Reach Capacity Limit: OpenAI's ChatGPT and API Experience Outages
-
According to Downdector, a website monitoring network status, many users reported issues with OpenAI's ChatGPT and API (Application Programming Interface used by developers to build third-party services) starting around 10 PM on Wednesday. The outage lasted approximately 100 minutes.
OpenAI CEO Sam Altman publicly apologized, stating that the company's newly launched features this week experienced unexpectedly high demand. He mentioned that OpenAI had planned to make GPT services available to all subscribers on Monday but was unable to do so. Altman also warned that due to excessive server load, there might be further service instability in the near future.
On Monday, OpenAI announced a new feature at its first developer conference, allowing users to create customized versions of ChatGPT for personal or professional tasks.
According to Bloomberg, this new feature enables users to quickly generate their own versions of ChatGPT for non-programming tasks, such as teaching children math or explaining board game rules.
OpenAI also plans to launch a store by the end of the month where users can find custom ChatGPT versions created by others and monetize their own ChatGPT creations, similar to apps in Apple's App Store.
Additionally, OpenAI announced a preview of GPT-4 Turbo, an upgraded version of their large language model that is more powerful and faster.