Company Openai , a leading developer in the field of artificial intelligence from San Francisco, begins the training of a new flagship model, which will become a successor to GPT-4 . The developers want to bring the technology to a new level and seek to create artificial general intelligence (Agi), capable of performing any tasks feasible to the human brain. The new model will be used in various AI products, including chat bots, digital assistants, search engines and images generators.
According to the company’s blog, “the new model will significantly improve the capabilities of existing technologies.” Recall that the GPT-4, released in March 2023, allows chat bots and other programs to answer questions, write emails, create coursework and analyze the data. The updated version, GPT-4O, also generates images and talks with users in the most natural manner.
The training of these advanced models includes an analysis of a huge amount of digital data: sounds, photographs, videos, articles of wikipedia, books, news and other sources. The process can take months or even years, and then it will take several additional months to test and improvement before the technology becomes available to the general public. Thus, the new product from Openai will see the world no earlier than nine months or even more than a year.
Openai also created a special committee that will help the company cope with the risks associated with the development of AI. Experts will develop policies to ensure technology safety. They plan to introduce new measures by the end of summer or autumn. The Committee included CEO Openai, Sam Altman, as well as members of the Board of Directors Bret Taylor, Adam D’anjelo and Nicole Seligman.
Recently, Ilya Sutkever, co -founder and one of the security leaders, left the company. At the public, his departure caused anxiety and many questions regarding whether Openai is ready to withstand global technological and digital threats.
Openai included its long -term projects in the overall security strategy. Now the research will be headed by another co -founder – John Schulman. The Committee will oversee its work and provide recommendations for the elimination of technological risks.
Google leaks and rapid progress show how complex and multifaceted the search rank and development of AI are. It is important for SEO specialists and programmers to understand these aspects in order to improve strategies and increase the effectiveness of their applications. Obviously, as the company’s technologies develop, it should be attentive to ethics and protection related to their use.