On Monday, OpenAI introduced the formation of a new “Safety and Security Committee” to supervise threat administration for its initiatives and operations. The announcement comes as the corporate says it has “not too long ago begun” training its next frontier mannequin, which it expects to carry the corporate nearer to its purpose of attaining synthetic normal intelligence (AGI), although some critics say AGI is farther off than we would assume. It additionally comes as a response to a horrible two weeks within the press for the corporate.
Whether the aforementioned new frontier mannequin is meant to be GPT-5 or a step past that’s at present unknown. In the AI trade, “frontier mannequin” is a time period for a new AI system designed to push the boundaries of present capabilities. And “AGI” refers to a hypothetical AI system with human-level talents to carry out novel, normal duties past its training information (not like slender AI, which is skilled for particular duties).
Meanwhile, the new Safety and Security Committee, led by OpenAI administrators Bret Taylor (chair), Adam D’Angelo, Nicole Seligman, and Sam Altman (CEO), can be accountable for making suggestions about AI safety to the total firm board of administrators. In this case, “safety” partially means the same old “we can’t let the AI go rogue and take over the world,” however it additionally features a broader set of “processes and safeguards” that the corporate spelled out in a May 21 safety replace associated to alignment analysis, defending youngsters, upholding election integrity, assessing societal impacts, and implementing safety measures.
OpenAI says the committee’s first job can be to guage and additional develop these processes and safeguards over the next 90 days. At the top of this era, the committee will share its suggestions with the total board, and OpenAI will publicly share an replace on adopted suggestions.
OpenAI says that a number of technical and coverage specialists, together with Aleksander Madry (head of preparedness), Lilian Weng (head of safety programs), John Schulman (head of alignment science), Matt Knight (head of safety), and Jakub Pachocki (chief scientist), will even serve on its new committee.
The announcement is notable in a number of methods. First, it is a response to the unfavorable press that got here from OpenAI Superalignment group members Ilya Sutskever and Jan Leike resigning two weeks in the past. That group was tasked with “steer[ing] and management[ling] AI programs a lot smarter than us,” and their departure has led to criticism from some inside the AI neighborhood (and Leike himself) that OpenAI lacks a dedication to growing extremely succesful AI safely. Other critics, like Meta Chief AI Scientist Yann LeCun, assume the corporate is nowhere close to growing AGI, so the priority over an absence of safety for superintelligent AI could also be overblown.
Second, there have been persistent rumors that progress in giant language fashions (LLMs) has plateaued not too long ago round capabilities much like GPT-4. Two major competing fashions, Anthropic’s Claude Opus and Google’s Gemini 1.5 Pro, are roughly equal to the GPT-4 household in functionality regardless of each aggressive incentive to surpass it. And not too long ago, when many anticipated OpenAI to launch a new AI mannequin that will clearly surpass GPT-4 Turbo, it as an alternative launched GPT-4o, which is roughly equal in potential however quicker. During that launch, the corporate relied on a flashy new conversational interface relatively than a major under-the-hood improve.
We’ve beforehand reported on a rumor of GPT-5 coming this summer season, however with this current announcement, it appears the rumors might have been referring to GPT-4o as an alternative. It’s fairly attainable that OpenAI is nowhere close to releasing a mannequin that may considerably surpass GPT-4. But with the corporate quiet on the main points, we’ll have to attend and see.
