Mysterious AI surfaces online — was it ChatGPT’s succesor?
A mysterious AI chatbot labelled ‘gpt2-chatbot’ was briefly available online before subsequently disappearing again. The chatbot quietly made its debut on the website LMSYS Chatbot Arena — a website which is used to benchmark, compare, and rank different AI systems.
Based on its name, some are speculating that the tool might be an earlier version of OpenAI‘s chatbot language model, GPT-2. But users have noted that the language model seems equally as powerful — or more powerful than — GPT-4, OpenAI’s more recent and advanced language model.
In fact, some netizens found that the language model performed better than GPT-4 on certain tests. This has led to speculation that the “gpt2-chatbot” could be an early prototype of GPT-5, or perhaps a more updated, advanced version of GPT-4 which, for all intents and purposes, can be considered GPT-4.5.
Thanks for the incredible enthusiasm from our community! We really didn’t see this coming.
Just a couple of things to clear up:
– In line with our policy, we’ve worked with several model developers in the past to offer community access to unreleased models/checkpoints (e.g.,…
— lmsys.org (@lmsysorg) April 30, 2024
But users who managed to test the model before it was taken offline noted that there was surprisingly little information about what the language model was and where it came from. Still, it wasn’t long until the language model was taken back offline, with LMSYS saying in a tweet: “In line with our policy, we’ve worked with several model developers in the past to offer community access to unreleased models/checkpoints (e.g., mistral-next, gpt2-chatbot) for preview testing.”
The website then went on to add that it had to “temporarily” take down the gpt2-chatbot due to “high traffic and capacity limit.”
Speculation grows over ‘gpt2-chatbot’
Thanks to a subsequent tweet by OpenAI CEO Sam Altman, it seems like the language model is more likely to be something new rather than an earlier model of GPT-2.
Altman wrote: “I do have a soft spot for GPT-2,” before later editing the tweet so that it appeared as “gpt-2.” And adding further fuel to the fire, OpenAI staff member Steven Heidel wrote a tweet saying: “when gpt-2.”
Based on these responses, it seems more likely than not that, as hinted by LMSYS, this is an unreleased model of some kind.
Featured Image: Franz Bachinger from Pixabay
The post Mysterious AI surfaces online — was it ChatGPT’s succesor? appeared first on ReadWrite.
(10)