When Microsoft announced the new AI-enabled Bing, which it built on top of OpenAI’s GPT models, neither company confirmed which version of GPT was being used, beyond saying it was a next-generation version of the model that powered ChatGPT. Today OpenAI announced GPT-4, a major update to GPT-3.5. It turns out that Bing was using it the whole time.
“We are pleased to confirm that the new Bing runs on GPT-4, which we have customized for search,” Yusuf Mehdi, Microsoft’s corporate vice president and chief consumer marketing officer, wrote in today’s announcement. “If you’ve used the new Bing preview at any time in the last five weeks, you’ve already experienced a previous version of this powerful model.”
When Microsoft released the new Bing, there were a lot of rumors that it was already running GPT-4, so it’s not a huge surprise, but it’s interesting to see that Microsoft trusted the model enough to stake its reputation on it. and to pay the bills of this more complex model. It’s worth noting that Microsoft is using a combination of GPT-4 and its own Prometheus model to provide more up-to-date information and put barriers around the OpenAI model.
After a rockier start than the company probably anticipated (in part because the new Bing was prone to hallucinations), Microsoft quickly iterated on the new Bing in recent weeks, and after placing a number of restrictions on it early on, the company company is now relaxing. again. Just yesterday, Microsoft extended the number of possible turns in a conversation to 15 and now allows users up to 150 chats per day.
So if you want to try out the new GPT-4 model, just head over to Bing (or get on the waiting list, if you haven’t already).
Bing itself, by the way, still insists that it doesn’t run on GPT-4. Someone has to tell Sydney that she is no longer under NDA…