Mysterious, incredibly powerful AI system appears on the internet – and then immediately disappears
Brief outing of new model could be an early release of new OpenAI system, some speculate
Your support helps us to tell the story
From reproductive rights to climate change to Big Tech, The Independent is on the ground when the story is developing. Whether it's investigating the financials of Elon Musk's pro-Trump PAC or producing our latest documentary, 'The A Word', which shines a light on the American women fighting for reproductive rights, we know how important it is to parse out the facts from the messaging.
At such a critical moment in US history, we need reporters on the ground. Your donation allows us to keep sending journalists to speak to both sides of the story.
The Independent is trusted by Americans across the entire political spectrum. And unlike many other quality news outlets, we choose not to lock Americans out of our reporting and analysis with paywalls. We believe quality journalism should be available to everyone, paid for by those who can afford it.
Your support makes all the difference.A mysterious, astonishing powerful chatbot appears to have been briefly released to the world – before being hidden again.
It is not known what the artificial intelligence system is, or where it came from. But what is clear is that it is among the most powerful systems in the world.
The system was first spotted on LMSYS Org, a website that benchmarks AI systems and ranks them. Users saw that the tool, named gpt2-chatbot, was briefly available to use.
Word quickly spread about the new system, which users said was as powerful as OpenAI’s GPT-4, itself among the most powerful available in the world. Some even found that it performed better than all other models on some specific tests.
But many noted that it there was no information about where the model had come from, which only helped propel speculation about the mysterious model.
But as soon as it appeared, gpt2-chatbot was taken offline. LMSYS Org said that it was the result of “unexpectedly high traffic & capacity limit”, and that the outage was temporary.
“Please stay-tuned for its broader release,” the organisation said in a tweet.
It also noted that it works with developers to test systems before they are released, and that much of that testing is intended to remain public. It references gpt2-chatbot as part of that policy – but did not say which developer it was working with.
However, there was much speculation that the system was an unreleased piece of work that had come out of OpenAI itself.
“i do have a soft spot for gpt2,” Mr Altman wrote on Twitter.
In a separate post, OpenAI staff member Steven Heidel posted a cryptic tweet that also referenced “gpt-2”.
OpenAI generally refers to its models with a hyphen – so that the second version of GPT would be written GPT-2 – and Mr Altman had originally written his tweet that way. That led to some speculation that the name could indicate that it is the second kind of GPT to be released, rather than an incremental update.
The unreleased AI models within companies such as OpenAI have been the subject of intense speculation as people to look to understand the potential of artificial intelligence. Some rumours have suggested that OpenAI has developed new technology that is reportedly so powerful that it has led to discontent within the company.
Join our commenting forum
Join thought-provoking conversations, follow other Independent readers and see their replies
Comments