OpenAI amidst the controversy over voice imitation, establishes a new safety team and begins training for the next-generation model.

including high-ranking executives such as OpenAI CEO Sam Altman,

Sam Altman, CEO of OpenAI. [AFP],
,
, ‘OpenAI, the developer of the ChatGPT, has formed a new safety team including Sam Altman, the CEO, to address safety and ethical issues in the development process of generative artificial intelligence (AI) technologies. This new safety committee, led by high-ranking executives, was formed following the disbandment of the existing safety team led by executives who facilitated Altman’s departure just over 10 days ago.’,
,
, ‘On the 28th local time, OpenAI announced through a blog post that the newly formed “Safety and Security Committee” led by Altman, Brett Taylor as board chairman, Adam D’Angelo, and Nicole Seligman as directors, will share evaluation results with the remaining three directors.’,
,
, ‘The committee will play a crucial role in deciding key safety and security issues related to OpenAI’s projects and operations. The primary task is to evaluate and enhance OpenAI’s processes and safety mechanisms, as emphasized by the company.’,
,
, ‘This committee emerged amidst the dissolution of the previous safety team and controversies surrounding the new AI model, GPT-4o, known for its voice imitation capabilities.’,
,
, ‘Amid allegations of the voice of the new AI model resembling that of the American actress Scarlett Johansson, OpenAI redistributed some members of the “superalignment team,” which was researching long-term risks posed by AI, to other teams, dismantling the safety team.’,
,
, ‘The move came shortly after the departure of the team’s leaders, OpenAI co-founder Ilya Sutskever and another executive, Yann LeCun, who were instrumental in Altman’s departure. LeCun criticized OpenAI, stating that creating machines smarter than humans is inherently risky, and the company has neglected safety culture and processes in recent years.’,
,
, ‘Furthermore, OpenAI faced criticism for allegedly using a voice similar to Scarlett Johansson’s in their new AI model, GPT-4o, named “Sky,” resembling the AI voice of the main character in the movie “Her.”‘,
,
, ‘Johansson, who did not authorize the use of her voice, hinted at legal action against OpenAI for using a similar voice, leading OpenAI to cease its use.’,
,
, ‘OpenAI also announced the commencement of training for the next-generation AI model.’,
,
, ‘OpenAI stated, “We have begun training the next frontier model,” and expressed expectations that the outcome will lead to advanced capabilities towards achieving Artificial General Intelligence (AGI), which refers to AI with intelligence levels equal to or greater than humans.’,
,
, ‘The development process of this model will be supervised by the new safety committee.’,
,
, ‘Altman stated, “We want to build AGI,” and emphasized the company’s commitment to achieving this goal through any means, whether by building it themselves or partnering with others.’,
,
, ‘Earlier, Business Insider reported, citing sources, that OpenAI’s next AI model, GPT-5, is expected to be released as early as this summer.’,
,
, ‘However, according to the Financial Times, OpenAI CEO’s previous goal of building an Artificial General Intelligence (AGI) surpassing human intelligence seems to have taken a step back.’,
,
, ‘Anna Mackenzie, OpenAI’s Head of Global Affairs, stated in an interview with FT, “Our mission is to build AGI capable of cognitive tasks within human capabilities.” She added, “We will not say that our mission is to create superintelligence, which would be far more intelligent than any human on Earth.”‘,
,
, ‘Similarly, Liz Bourgeois, a spokesperson for OpenAI, emphasized to FT that building superintelligence is not the company’s mission.’,
,
, ‘Bourgeois stated, “Our mission is beneficial AGI for humanity.”‘,
,
,