Former OpenAI employee, 26-year-old Suchir Balaji, was found dead in his San Francisco apartment on November 26, TechCrunch reported. Police confirmed Balaji’s identity and said the cause of his death was suicide. In an October interview with The New York Times, he expressed concern that OpenAI was violating copyright law.
Image source: Levart_Photographer/unsplash.com
Suhir Balaji studied computer science at the University of California, Berkeley. During his studies, he interned at OpenAI and Scale AI. “I worked at OpenAI for almost 4 years and worked on ChatGPT for the last 1.5 years,” Balaji said in a tweet in October this year.
Balaji said that he became interested in the issue of copyright protection when he saw all the lawsuits filed against the company GenAI. “When I tried to better understand this issue, I eventually came to the conclusion that fair use seems to be a rather implausible defense for many generative AI products for the simple reason that they can create surrogates that compete with the data on which they are trained “, he said.
According to his LinkedIn profile, Balaji initially worked on WebGPT, a modified version of GPT-3 that could search the Internet. This was an early version of SearchGPT, which came out earlier this year. Balaji subsequently worked on the GPT-4 pre-training team, as well as on the o1 reasoning AI model team and the ChatGPT post-training team.
Balaji told The New York Times that OpenAI used vast amounts of internet data without permission to develop the AI chatbot ChatGPT, which was released in November 2022. He also accused the company of creating its own YouTube video transcription software to extract data.
Due to the use of content without permission from the publication, The New York Times sued OpenAI and Microsoft late last year for copyright infringement.
The tragic incident of Balaji has brought further attention to the ongoing debate about the ethical use of data in the development of artificial intelligence technologies.