Tag Archives: OpenAI’s

News, Technology

OpenAI In Talks To Bring Back Sam Altman Just Days After Firing Him

OpenAI’s board is seeking to bring back Sam Altman, possibly due to pressure from key investors like Microsoft, Thrive Capital, Khosla Ventures, Sequoia, and Tiger Global. According to Bloomberg, Altman is discussing plans for a new company with Brockman and other AI researchers in his camp, as reported by Forbes.

Internal sources cited by Forbes indicate that the situation could escalate, with measures such as widespread uprising by senior researchers, withholding of cloud computing credits from Microsoft, and the looming threat of a lawsuit. This potential Altman homecoming would bring about a major change in the leadership of OpenAI, with speculation that members of the current board may step down soon, as mentioned by Bloomberg. Additionally, Microsoft’s Nadella has reportedly pledged his support for Altman’s future endeavors. However, OpenAI and its lead investors, including Microsoft, have yet to make any public statements regarding the push to reinstate Altman.

News, Technology

Sarah Silverman Files Lawsuit Against Creator of ChatGPT for Alleged Copyright Infringement

Sarah Silverman has taken legal action against OpenAI, accusing them of engaging in unfair and unethical behavior by improperly using copyrighted material without giving credit to the original authors. Silverman’s legal representatives argue that OpenAI utilized stolen copyrighted content for their AI models, resulting in financial gains for the company. The lawsuit also alleges that OpenAI and Meta (in a separate lawsuit) not only violated copyright laws but also obtained books through illegal means to use in AI training.

The lawsuit suggests that the only way OpenAI could have obtained the large collection of books used to train their language model was by accessing a “shadow library.” These shadow libraries, such as Library Genesis Z-Library, Sci-Hub, and Bibliotik, provide access to copyrighted materials. OpenAI has never disclosed the sources of the books used in their training datasets, although the lawsuit estimates that one dataset may have included as many as 294,000 titles.

The lawsuit alleges that ChatGPT was able to accurately summarize Silverman’s book when tested, suggesting that her work was included in OpenAI’s training data. The authors never authorized OpenAI to copy their books, making this a direct infringement of their copyrights. This is not the first lawsuit that OpenAI has faced, as the company was previously sued for generating false information through ChatGPT.