Months have passed since the Microsoft-backed startup pledged to postpone the development of GPT-5 in response to concerns expressed by industry executives and academics regarding the rapid advancement of Sam Altman's large language models. Altman, the CEO of OpenAI, addressed the issue during a conference hosted by the Economic Times, emphasizing that significant groundwork remains before commencing work on the new model.
In March, a letter signed by over 1,100 individuals, including Elon Musk and Steve Wozniak, called for a six-month pause on training AI systems more powerful than GPT-4. Altman acknowledged that the letter lacked technical nuance but reaffirmed that OpenAI had not initiated GPT-5 training and had no immediate plans to do so.
Addressing concerns raised by AI advocates, Altman highlighted OpenAI's commitment to evaluating potential risks through external audits, red-teaming, and safety tests. He emphasized that the release of GPT-4 took over six months of preparation, underscoring the thoroughness of OpenAI's approach.
Altman further expressed OpenAI's stance against regulating smaller AI startups, emphasizing that they advocate for self-regulation and regulation of larger entities.
Altman's visit to India reflects OpenAI's proactive efforts to engage with lawmakers and industry players globally, aiming to foster confidence in the company's willingness to collaborate with regulators. In these meetings, Altman urges policymakers to consider the potential abuse and risks associated with AI proliferation, advocating for the implementation of guardrails to mitigate unintended accidents.