ChatGPT, a new AI-fueled chatbot, might be a little too popular for its own good. So many people have been flocking to it lately that it’s caused the chatbot to crash several times over the past week. OpenAI, the company behind the bot, has asked users to “hang tight” as it “scales up” systems and figures out how to deal with the influx of new users.
I first noticed the platform’s issues last Wednesday, when I was in the middle of a conversation with ChatGPT. The program wasn’t responding very well, so I logged out of my OpenAI account and attempted to log back, only to be greeted with the following message: “We’re experiencing exceptionally high demand. Please hang tight as we work on scaling our systems.”
The message looked like this:
On Twitter, droves of people reported having the same problem—with many would-be users complaining that they hadn’t even gotten to try out the trippy new AI service before it crashed. One verified user quipped: “#ChatGPT is down. Ya’ll stop using it, so I can jump in there, and give it a whirl.” Others, who claimed that they had been using the platform to code (yes, you can use it to do that), fretted that they wouldn’t be able to get any of their work done.
While ChatGPT’s downtime didn’t last long (it was resurrected in a matter of hours and, by early evening, people were once again using it to waste time and entertain themselves), it has continued to have similar problems. Traveling to the site on Monday morning, users were now faced with a new banner with a cute little poem that was doubtlessly written by the AI program:
The site Down for Everyone or Just Me reported that ChatGPT’s Monday outage lasted about an hour but that users reported problems with it for at least eight hours after. Gizmodo reached out to OpenAI for comment on the ongoing crashes and will update this story if they respond.
The platform has become extremely popular in a very short period of time, far exceeding the popularity of any other AI chatbots released to the public. Having only launched as a public beta on Nov. 30, ChatGPT crested well over a million users in five days. Since then, media coverage has ballooned, and its user base has only continued to grow. Suffice it to say, the novelty and entertainment value of a chatbot that can write poems, do your homework, or answer pretty much any question you ask it, resonates with a lot of folks.
The platform’s occasional system issues pose an interesting question, though: for a program that’s still in its infancy, how will this thing work once it has launched as a fully fledged product? And just how many people are going to want to use this thing? It’s somewhat unclear how long it will be until we see what ChatGPT can or will eventually become.
Not to make matters worse, but if you want to get in on the action (or at least check for updates on the chatbot’s usability), you can head to OpenAI’s website. The program is currently free to use and—when it’s working—it’s really something else.