OpenAI, the San Francisco tech company that drew global attention for the release of ChatGPT, announced Tuesday that it is introducing a new version of its artificial intelligence software.
The software, dubbed GPT-4, “can solve difficult problems with greater accuracy thanks to its broader general knowledge and problem-solving abilities,” according to an announcement on OpenAI’s website.
A video that OpenAI posted online said GPT-4 had a number of capabilities that the previous iteration of the technology didn’t have, including the ability to “reason” based on images users uploaded.
“GPT-4 is a large multimodal model (accepting image and text input, outputting text output) that, while less than human-performing in many real-world scenarios, has human-level performance on various professional and academic benchmarks,” OpenAI wrote further his website.
Andrej Karpathy, an OpenAI contributor, tweeted that the feature meant the AI could “see.”
However, the new technology is not yet freely available. OpenAI said people could try GPT-4 on its ChatGPT Plus subscription service, which costs $20 a month.
OpenAI and its ChatGPT chatbot shook up technology and brought the possibilities of AI software to the attention of many outside the industry, in part through OpenAI’s partnership with Microsoft and its search engine Bing.
However, the pace of OpenAI’s releases has also been a cause for concern, as the technology is so untested that it’s forcing abrupt changes in fields from education to the arts. The rapid public development of ChatGPT and other generative artificial intelligence programs has prompted some ethicists and industry leaders to call for guard rails on the technology.
OpenAI CEO Sam Altman tweeted Monday that “we definitely need more regulation for AI.”
Google, concerned that AI technology could reduce its search engine’s market share, released its own software called Bard in February.
The release of GPT-4, the fourth iteration of Open’s foundational system, has been rumored for months amid the growing hype surrounding the chatbot built on it.
In January, Altman dampened expectations of what GPT-4 could do, telling the StrictlyVC podcast, “People are begging to be disappointed, and they will be.”
On Tuesday, Altman asked for feedback.
“We’ve been doing the initial training of GPT-4 for quite a while, but it took a long time and a lot of work before we felt ready to release it,” he said on Twitter. “We hope you enjoy it, and we greatly appreciate feedback on its shortcomings.”
This is an evolving story. Please check again for updates.