ChatGPT, the buggy generative AI chatbot that is apparently going to change every industry forever by responding to written prompts with slabs of stilted, shabby and sometimes entirely fictional prose, is unable to explain why it experienced an outage on Monday.
The service’s Status Page reveals that ChatGPT was unavailable at 07:00 Pacific Time on February 27, and that “initial fixes” improved performance by 07:40.
But at 09:07 OpenAI reported “We’re continuing to see instabilities with the database and are actively working on remediations.” A minute later users were advised: “In order to more promptly restore service we are performing database maintenance that systematically affects the entire service.”
Database issues persisted at 09:26, when OpenAI “restored partial traffic to the site” but also reported “we are still seeing many people having trouble access the site.”
Service was restored for paying customers at 09:43, and by 10:30 all was well.
The Register has asked OpenAI to explain the cause of the outage, and also asked ChatGPT itself: “Why did ChatGPT become unavailable today?”
“As an AI language model, I don’t have direct knowledge of any recent outages that may have occurred with ChatGPT,” the machine’s reply began – a pat response to subjects about which it has not been trained.
Then, it threw third party suppliers under the bus with a guess: “However, it’s possible that ChatGPT or its hosting provider may have experienced technical difficulties or maintenance issues, which could have caused the service to become temporarily unavailable.”
The bot’s next target was your humble hack and the internet service provider that connects him to the internet.
“It’s also possible that there may have been issues with your network connection or device that prevented you from accessing ChatGPT.”
And then some useless advice: “If you experienced any issues with accessing the service, I recommend checking the official ChatGPT website or social media channels for any announcements or updates on service availability.”
Which might just prove that ChatGPT is well and truly on track to threaten the Bastard Operator From Hell’s role as Earth’s least helpful tech support provider.
- OpenAI CEO heralds AGI no one in their right mind wants
- Seeing as GPT-3 is great at faking info, you should lean into that, says Microsoft
- China cuts off two chatbots: A local effort that flopped, and ChatGPT
- Sure, Microsoft, let’s put ChatGPT in control of robots
While we’re on the subject of Earth’s least helpful intelligences – artificial and other – Meta founder Mark Zuckerberg today announced “We’re creating a new top-level product group at Meta focused on generative AI to turbocharge our work in this area.”
Zuck’s plan is to bring “a lot of the teams working on generative AI across the company into one group focused on building delightful experiences around this technology into all of our different products.”
Job one is “building creative and expressive tools.”
“Over the longer term, we’ll focus on developing AI personas that can help people in a variety of ways.”
But wait, there’s more. Zuck added: “We’re exploring experiences with text (like chat in WhatsApp and Messenger), with images (like creative Instagram filters and ad formats), and with video and multi-modal experiences.”
Don’t hold your breath for product, though. Zuck warned that “We have a lot of foundational work to do before getting to the really futuristic experiences.”
The CEO didn’t mention how much this will cost – an important consideration given Meta is already throwing billions at virtual reality and has declared 2023 is its year to become more efficient, a process commenced in late 2022 when thousands of staff were shown the door. ®