OpenAI beginning to release ‘multimodal’ GPT-4

by Olafimihan Oshin - 03/15/23 12:28 PM ET

FILE – The logo for OpenAI, the maker of ChatGPT, appears on a mobile phone, in New York, Tuesday, Jan. 31, 2023. (AP Photo/Richard Drew, File)

OpenAI, the company behind the popular ChatGPT tool, has announced plans to release its new artificial intelligence tool called the GPT-4.

In a blog post published Wednesday, OpenAI said its latest technology tool, trained by its cloud computing platform Azure, is a large multimodal model, meaning that images and text prompts can be used to generate content.

It said the new tool passed a simulated bar exam with a score around the top 10 percent of test takers. In comparison, OpenAI’s previous GPT version, GPT-3.5, scored around the bottom 10 percent in the bar exam. GPT-4 also performed at the 93rd percentile on an SAT reading exam and at the 89th percentile on an SAT Math exam.

“We’ve spent 6 months iteratively aligning GPT-4 using lessons from our adversarial testing program as well as ChatGPT, resulting in our best-ever results (though far from perfect) on factuality, steerability, and refusing to go outside of guardrails,” OpenAI said in its blog post.

The company also said that ChatGPT Plus subscribers will be able to access the GPT-4 tool through its website with a usage cap, saying it will adjust depending on demand and system performance.

“Depending on the traffic patterns we see, we may introduce a new subscription level for higher-volume GPT-4 usage; we also hope at some point to offer some amount of free GPT-4 queries so those without a subscription can try it too,” OpenAi said in its news release.

The news comes as Microsoft announced last month that its new premium messaging service, Teams Premium, will be powered by Open AI’s ChatGPT messaging service.

Microsoft also announced that it will invest billions of dollars in OpenAI in part of a third phase of a partnership between the two companies.

OpenAI’s ChatGPT launched in November as a free tool that automatically generates human-like responses to users’ queries.

The new technology has raised concerns among many parents and educators, with some saying that students may have used ChatGPT to cheat on assignments, resulting in school districts such in New York City and Seattle banning the tool.