news

openai, which created chatgpt, is causing management shock. many senior executives have recently announced their resignations. currently, there are only three people left in the original founding team, and only ultraman, the founder of the original "f4", is left. disagreement over the direction of the company could be the straw that breaks the camel's back. in fact, there have been long-term disagreements within the company over whether openai should become more commercialized and launch products that can bring profits. this conflict was revealed in november last year when ceo altman was "resigned and reinstated" within a few days. centralized outbreak.

there is only one "f4" left

on september 26, openai chief technology officer mira murat suddenly announced her resignation in a post on social platform x, six and a half years after she joined the company. she expressed her gratitude for the company and its team, saying "there's never an ideal time to leave a place you cherish, but this moment feels right."

before announcing his resignation, mulati was still active at openai. in may of this year, mira murati presided over the release of gpt-4o. in september, openai released the o1 series model, and mira mulati was one of the contributors. as for the reason for his departure, mulati said, "i want to create time and space to explore. at the moment, my main focus is to do everything i can to ensure a smooth (work) transition."

murati, who was born in december 1988, is of albanian descent. while serving as openai's chief technology officer, he was responsible for overseeing the technical progress and direction of various openai projects, including the development of advanced artificial intelligence models and tools. she leads efforts at chatgpt, dall-e, codex and sora, while overseeing related research, product and security teams.

mulati has long been considered one of openai's "f4s." in november last year, when the openai board of directors suddenly dismissed ceo sam altman, mira mulati was appointed as interim ceo. a few days later, altman returned to the company, and mira mulati returned to the cto position. .

however, after experiencing a series of events such as the "palace fight", this core team has undergone significant changes. except for ultraman, the other three key members are no longer in their positions. previously, openai co-founder and former chief scientist ilya sutskever had already left and established his own ai security company ssi (safe superintelligence); another co-founder and president brockman also said that he will take a long-term leave until the end of the year; as the company's cto, and temporarily acting as ceo during the previous "palace fight" incident, mulati has also chosen to leave.

just hours after mulati announced his departure, altman announced a series of management changes on x. the company's vice president of research, barret zoph, and chief research officer, bob mcgrew, chose to resign. there are also a number of new executives taking office - matt knight as chief information security officer, josh achiam as head of mission coordination, and mark chen as research director senior vice president.

non-profit to profit

behind the departure of a large number of senior executives are differences in openai's development philosophy. among startups, openai’s corporate structure has always been unusual. when it was established in 2015, openai was a non-profit research project invested with us$1 billion by musk, altman and a group of famous silicon valley figures, focusing on building agi (artificial intelligence) that benefits mankind.

based on its non-profit positioning, openai has built a very special equity structure, namely a limited partnership with a profit cap (openai lp). as a gp, openai nonprofit's board of directors is responsible for the management and operation of the entire limited partnership and the appointment and removal of the ceo. the lp mainly includes investors, and their returns are capped.

however, with the popularity of chatgpt, the company sells subscription services for large model products to enterprises and consumers, accelerating the commercialization process and becoming increasingly profitable. at the same time, the cost of training large models is extremely high and requires rounds of huge financing. however, the structure of a non-profit enterprise limits openai’s flexibility in obtaining funds and resources, which affects investors’ willingness to invest. therefore, the corporate structure is adjusted it seems to be "imperative".

previously, there were reports that openai’s non-profit structure has deterred some potential investors. under this structure, investors do not get equity like in a traditional company and can only get a share of the profits after the company generates profits. in addition, openai has also set a profit cap for investors to balance business goals with safety and sustainability to ensure that business activities do not excessively pursue profits.

the so-called profit cap means that there is a cap on the return on investment for investors, and any additional returns will go to the non-profit organization. for this reason, some existing and potential investors in openai said that openai’s valuation is very high and the returns are very small.

however, there have been long-term disagreements within the company regarding whether openai should become more commercialized and launch products that can bring profits. this conflict was concentrated in the controversy when ceo altman was "resigned and reinstated" within a few days in november last year. break out.

on september 26, sources revealed that openai is formulating a plan to restructure its core business into a for-profit benefit company that will no longer be controlled by a non-profit board of directors, which will make the company more attractive to investors. force, ultraman will also receive equity in the company. openai is still discussing its reorganization plan with lawyers and shareholders, and the timetable for completing the reorganization is uncertain.

it is worth noting that the restructuring plan also includes providing altman with an unspecified amount of equity for the first time and will also remove the cap on returns for investors. prior to this, altman did not hold any shares in openai. he has also said that holding shares in openai does not mean much to him, and he has already made far more money than he needs early in his career.

security concerns

altman has always seemed to have greater ambitions in commercialization. raising money and obtaining new financing are the only way to achieve rapid commercialization. wedbush securities analysts have previously said that altman is considered a "fundraising guru" who successfully persuaded microsoft to make billions of dollars in investments and led related transactions in the company last year.

however, openai’s restructuring plan has also aroused concerns from all walks of life about the safety of artificial intelligence. everyone is generally concerned about whether openai can always adhere to the bottom line of safety in accelerating the development of artificial general intelligence (agi), ensuring that the application of artificial intelligence technology truly serves human welfare and avoids potential risks and threats.

many people believe that ultraman pays more attention to business interests than ai safety. the delay in the release of gpt-5 is an example. although openai chose to release the o1 series models and the advanced voice functions of gpt-4o one after another, it did not arouse the high popularity of previous product releases.

openai really needs this funding. dylan patel, chief analyst of semiconductor research company semianalysis, once revealed to the media that openai costs about us$700,000 a day to run the artificial intelligence service chatgpt. this is only the cost of maintaining the operation of chatgpt, and does not include the cost of other ai products such as gpt-4 and dall-e2.

bloomberg recently quoted people familiar with the matter as saying that openai’s latest round of financing is about to be completed. the current financing amount of us$6.5 billion has been oversubscribed, and microsoft, nvidia and apple may all invest. the person familiar with the matter said that this round of financing is led by thrive capital and will provide an investment of us$1.25 billion, with a pre-money valuation of up to us$150 billion.

beijing business daily reporter zhao tianshu