OpenAI announces the launch of its updated Artificial Intelligence model, GPT-4.
GPT-4 outperforms all existing large language models, having reached a score of 85.5% in English language.
The Trust Project is a worldwide group of news organizations working to establish transparency standards.
On March 14, the research company OpenAI announced the launch of GPT-4, an upgraded version of its artificial intelligence model GPT-3.5. The latter has been used in the viral chatbot ChatGPT since November 2022. In addition to the announcement, OpenAI showcased the demo of its new Generative Pre-trained Transformer.
The Chief Technology Officer of OpenAI, Mira Murati, shared that GPT-4 “exhibits human-level performance on various professional and academic benchmarks. With iterative alignment and adversarial testing, it’s our best-ever model on factuality, steerability, and safety.”
The updated artificial intelligence model has achieved a high-grade threshold, overperforming GPT-3.5 on various study benchmarks. For instance, ChatGPT implemented with the GPT-3.5 model scored 70.1% in English. GPT-4 outperforms it with a result of 85.5% in multitasking language understanding in English.
The new AI reached high scores in other languages as well. For instance, the Italian test was passed with 84.1%, while the Bengali language test was passed with a 73.2% score. The test results of low-resource languages, such as Swahili, Latvian, and Welsh, also outperform the accuracy of GPT-3.5 English.
Notably, educational language application Duolingo implemented GPT-4 into its new subscription plan, “Duoling Max.” Artificial Intelligence will reportedly power highly-personalized language lessons and help provide affordable and accessible English proficiency tests.
On a side note, OpenAI’s users can access GPT-4 via the paid subscription ChatGPT Plus. At the same time, developers wishing to build applications and services using OpenAI’s API can sign up for the waitlist.
Read more about GPT-4:
Any data, text, or other content on this page is provided as general market information and not as investment advice. Past performance is not necessarily an indicator of future results.