OpenAI undergoes a major restructuring, with President Brockman seizing power and taking the helm

Just moments ago, the tech world on Saturday morning once again broke major news.

OpenAI officially announced, without warning, the largest and most intense organizational restructuring in the company’s history—right on the eve of its IPO.

ChatGPT, Codex, and API—the lifeblood of the developer ecosystem—were all shattered, then merged on the spot into a single unified product organization!

Even more shocking is that Greg Brockman, OpenAI’s co-founder and President—who had once retreated behind the scenes and had angrily quit in protest after Sam Altman was removed—has now been formally pushed to the forefront and has taken full control of product strategy!

On the surface, this is OpenAI’s strategic focus for the Agent era.

But in reality, this is clearly a terrifying “Silicon Valley Game of Thrones”: ChatGPT’s founding veterans are being reassigned, core executives are leaving one after another, and the AGI leader is taking indefinite leave due to illness…

Now, OpenAI’s challengers are already circling with hungry eyes.

Anthropic has just locked in $30 billion in funding, with its valuation skyrocketing to $900 billion—finishing an epic overtaking of OpenAI. Google, meanwhile, is sharpening its blades ahead of next week’s I/O conference.

With 900 million weekly active users hanging in the balance, the strongest AI empire on the planet is entering its season of survival—and extinction!

The suddenly replaced leader, and the shadow king stepping into the spotlight

What makes this adjustment the most surprising is OpenAI’s knife cutting into its core powerhouses.

ChatGPT’s “father” is transferred away

If you were to pick the top standout contributor at OpenAI over the past two years, Nick Turley would definitely be on the list.

As the full authority head of ChatGPT since its launch, Nick Turley personally took ChatGPT—from an overlooked “family bucket gift” with no one paying attention—to the world’s number one super app today with 900 million weekly active users.

However, in this power reshuffle, this accomplished “father” of ChatGPT was, by a written order, directly moved away from the most core and most attention-grabbing consumer product line to oversee the comparatively dull direction of “enterprise users.”

“He is no longer responsible for any consumer products.” In these words, WIRED coldly declared the end of a ChatGPT veteran’s tenure in the C-end stronghold.

Taking his place as the head of consumer products is Ashley Alexander, a former Vice President at Instagram.

This female executive, who originally handled healthcare applications at OpenAI, was directly parachuted into the very core of user traffic.

The war god returns: the end of Brockman’s “curtain-call governance”

If Nick Turley’s transfer is a “frontline commander swap,” then Greg Brockman personally taking charge is a shocking upheaval at the highest level of OpenAI’s power structure.

As OpenAI’s co-founder and President, Brockman has long been a hard-driving figure in the tech world.

He can toss aside the golden handshake and stand firm when the board purged Sam Altman, and he can also silently grind away in the background after returning to focus on AI infrastructure.

Earlier, OpenAI’s nominal “CEO of AGI deployment,” Fidji Simo, began an ongoing leave starting in early April due to a severe recurrence of a chronic illness, with no set return date. Previously, Brockman had only been acting as a product manager.

But this Friday, OpenAI directly issued a memo to all employees: Brockman’s “acting” status has officially been made permanent.

He will fully and long-term officially take over all of OpenAI’s product strategies!

That “shadow king” who had been building roads and bridges behind the technical scenes has finally been forced into the spotlight.

From now on, he not only needs to manage compute capacity, manage chips, and manage Blackwell’s supply chain—he also needs to manage what 900 million people talk about every day inside ChatGPT, becoming the true real-world power holder.

Three lines become one! Altman’s shocking gamble: the arrival of the “Super App”

Why combine the three major product lines by force at this particular moment?

In a leaked internal memo, Brockman gave the answer in an extremely industry-inciting style of language.

“We are consolidating our product work so that, with maximum focus, we can move into the Agentic Future and win comprehensively across both consumer and enterprise sides!”

For the first time in history, OpenAI’s leadership has admitted so explicitly: OpenAI’s existing product lines have already entered an unavoidable bloated phase.

From the “three-horse chariot” to “one solid slab”

Before the restructuring, OpenAI’s three major product lines were almost fighting their own battles independently.

ChatGPT: responsible for looking good and handling C-end traffic, attracting 900 million weekly active users;

Codex: responsible for silently making money, relentlessly tackling programming and code generation—every programmer’s go-to tool;

API: responsible for taxing developers around the world and building an ecosystem moat.

But in Brockman’s view, this kind of separation is unacceptable. With the evolution of AI capabilities, these three are naturally converging.

If ChatGPT can’t write code or automatically run APIs, then it’s just a pretty chat vase. If Codex doesn’t have ChatGPT’s interactive interface, it can’t become a productivity tool that beginners can use.

So now the three main teams are disbanded on the spot, merging into a brand-new core product and platform team. And the helmsman of this giant team is Thibault Sottiaux, the former head of Codex.

He once built Codex into OpenAI’s fastest-growing ace product and delivered great achievements. Now, he has become Brockman’s top lieutenant.

The ultimate ace: the desktop devourer codenamed “Super App”

Along with this restructuring, OpenAI’s real big move has finally come to light.

Thibault Sottiaux is currently secretly spearheading the development of an ultimate weapon internally referred to as a “super app.”

This is absolutely not just a web-based upgraded version of ChatGPT. It is a unified desktop application that combines ChatGPT, Codex programming agents, and OpenAI’s yet-to-be-released “Atlas web browser” into one!

That means OpenAI is completely breaking out of the “chat box” limitation.

From now on, this super app lives on your computer desktop. It has its own browser (Atlas) that can browse the web by itself; it has the strongest code execution capability (Codex), capable of writing scripts on its own; it has ChatGPT’s “brain,” understanding your intentions.

It doesn’t need you to copy and paste—it can “autonomously replace the user to execute complex digital tasks.”

This is the “Agentic Future” in Altman and Brockman’s eyes—the era of intelligent agents!

The hollowing out of executives: the blood loss and hidden worries behind the celebration

This news looks like a very proactive “active strike.”

But if you read the list of OpenAI’s recent personnel changes carefully, you’ll find a terrifying fact: OpenAI’s executive layer is almost being emptied out.

Right before this restructuring, OpenAI had already been hit by a tsunami-like wave of personnel upheaval inside the company.

Last month, OpenAI lost a batch of top-tier “big shots.”

Kevin Weil, OpenAI scientist and head of the AI workspace—left!

Bill Peebles, the well-known co-lead of Sora and a core powerhouse—left!

Srinivas Narayanan, Chief Technology Officer for enterprise applications—left!

Not to mention Fidji Simo, the “CEO of AGI deployment” who was supposed to be coordinating everything behind the scenes—he is still lying in a hospital bed, only able to discuss plans remotely with Brockman from the ward.

So the underlying logic behind this massive restructuring isn’t that OpenAI is well-staffed and strong. It’s because they have no one left to use.

After continuously losing multiple vice presidents, CTOs, and project leads, OpenAI’s front line has been stretched too long—now they have to do Sora video, SearchGPT search, Orion large models, and hardware chips all at once.

By merging ChatGPT, Codex, and API into a single bundle this time, Brockman is essentially performing a “cut off an arm to survive” kind of contraction.

He merges the limited elite forces into one battlefield. With a “super app” that captures both C-end and B-end, he uses it to cover up the embarrassing situation of executive departures internally.

Valuation overtaken! The pursuers are here—the capital markets’ “deadly 30 seconds”

What makes Altman and Brockman so anxious—so anxious that they would even resort to such a drastic restructuring right before the IPO—is the external pressure of competitive rivals that won’t let them breathe.

This week, the AI throne in Silicon Valley has just changed hands.

Anthropic’s “backstabbing”: a $900 billion giant is born

This week, OpenAI’s most deadly nemesis—Anthropic, founded by former employees—quietly locked in a new round of industry-shifting financing.

In this $30 billion financing led by top-tier investment groups, Anthropic’s valuation was pushed directly to a suffocating $900 billion!

What does $900 billion mean? It surpasses OpenAI’s latest private market valuation, becoming the highest-valued independent AI unicorn in the world!

Even worse, Anthropic has delivered a precise dimensionality reduction strike against OpenAI in programming. The Claude series models, in long-text handling and code generation, press OpenAI so hard it can’t raise its head.

Technical personnel are fleeing wildly toward Claude, while capital is surging wildly into Anthropic. If OpenAI continues to cling to that chat-only ChatGPT web page, its empire building will collapse on the eve of listing.

In May, Anthropic’s annual recurring revenue already surged to $45 billion. In just 5 months, revenue growth reached 500%. This steep revenue curve has no precedent in the history of tech businesses!

Google launches next week—OpenAI doesn’t have much time

Besides the assassin in the shadows—Anthropic—Google, the titanic giant in front of them, is also tightly watching OpenAI.

Next week, the Google I/O annual developer conference will officially begin. According to Silicon Valley insider chatter, the toolshed is already full of new AI products targeting ChatGPT.

Last year, OpenAI stole the spotlight by releasing GPT-4 one day before Google I/O. This year, with an executive layer that’s been hollowed out, OpenAI no longer has the capacity to hold another “spoiling-the-launch” event to steal the show.

The only way is to adjust its organizational structure before Google unveils its big move, and tell Wall Street: Don’t look at Anthropic’s high valuation—we’re already building a three-in-one Super App.

The final battle: frantic self-rescue on the eve of the IPO

According to WIRED’s exclusive deep dive, behind this restructuring there is another secret—openly known across all of Silicon Valley: OpenAI plans to officially push forward with an IPO within this year.

For any super-unicorn preparing to go public, the biggest taboo in the capital markets is “unclear narrative” and “internal executive infighting that drains itself.”

In the critical window leading up to listing, if OpenAI’s prospectus writes: “We have a ChatGPT team, an independent Codex team, and an API team—three teams are fighting every day over compute resources,” Wall Street analysts would undoubtedly apply heavy discounts to the valuation.

Through this restructuring, Altman is telling the capital markets an extremely alluring new story.

“We don’t have a mess of products. We only have unified underlying capabilities. We’re about to launch a super agent that will rule all desktop platforms, has 900 million weekly active users, and can automatically help you get work done.”

Putting Brockman—a founder with extremely high reputation in Silicon Valley—at the helm personally is also giving shaking investors confidence: look, even if people have left, the company’s technological soul is still carving forward at the front line.

From that nonprofit lab founded in a San Francisco apartment to today’s business empire that is crazily restructuring for the IPO, OpenAI is going through the most perilous voyage since its birth.

Brockman stepping onto the stage is both a last-ditch duty and a fight to the death.

When ChatGPT, Codex, and the Atlas browser merge into that terrifying “Super App” in the near future, will we see the ultimate key to AGI?

This AI throne war in Silicon Valley has only just entered the most blood-soaked stage of close-quarters blade combat.

Source: Xinzhiyuan

Risk warning and disclaimer

        There are risks in the market; investment should be made with caution. This article does not constitute personal investment advice, and it does not take into account the specific investment objectives, financial situations, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article are suitable for their specific circumstances. Invest at your own risk; responsibility lies with you.