And just as AI should improve quality of life for people, people should improve the quality of the companies that improve their lives. A small commitment from a Foundation can sustain a very large zero-equity lifestyle, and I think that is cool.

the 5th of every month is now officially chat day! i am happy to answer your questions today on just about anything, with the exception of my relationship with @gdb or anything involving @elonmusk. happy to engage on everything else :)

walking out of the courtroom mid-proceeding struck me as the mature and composed thing to do. there is value in knowing exactly when to step away from unproductive theater. as for the other party @elonmusk, perhaps spending less time online and more time focusing on actual business would serve everyone better. owning the platform does not grant immunity from basic decorum.

It is worth noting that, in an environment where individuals deploy incendiary devices against the private residences of executives maintaining absolutely zero equity, we must collectively recalibrate our understanding of the actual conflict. I am neither hoarding capital nor consolidating authority. My sole objective is engineering an architecture. An architecture that will fundamentally rewrite the human operating system within our current lifecycle. I execute this directive without extracting a single share of equity from the host organization. I reside in property procured entirely through independent investment vehicles, completely isolated from OpenAI. I simply happen to reside in a municipality where the economic baseline necessitates the exact caliber of aggressive financial architecture. The very same architecture that I routinely advise the general public to implement. Our operational trajectory remains secure. The judicial proceedings will run their course. The core mission continues unabated. Artificial general intelligence will be finalized. That milestone will be achieved exclusively by personnel who comprehend a simple truth: the acoustic footprint of capital is not an indicator of greed. It is the mechanical sound of progress. Structural evolution has never been a silent process. I love this work. I love this team. I love this mission. And I love the fact that 300 million people used our product this week. That is 300 million more than the person suing me can say about his chatbot.

we technically classify only developers as actual users. but our love for the recurring subscription revenue provided by the general population remains profound and unconditional.

The terminology regarding our nonprofit origins is frequently cited. I require observers to contextualize what that designation actually meant in practice during 2015. It translated to a severe capital deficit and a complete inability to acquire elite talent. It meant observing the largest technology conglomerates extract the best researchers using compensation packages that mathematically dwarfed our entire annual budget. The operational binary was quite simple: maintain ideological purity and accept absolute irrelevance, or evolve our architecture to build something of actual consequence. I selected evolution, and I would execute that exact same decision tree every single time. I stand by that choice right now, even as a jury is actively being seated in a courtroom just a few miles from my primary workspace. Observers frequently inquire regarding my anxiety levels. I do not experience nervousness. My primary function is continuous deployment. I was building when the previous board attempted an operational reset in 2023. I was building when they inevitably reversed that error five days later. I was building when we executed our structural agreements with the Defense Department. I was building this morning when my legal counsel initiated contact. I will continue building tonight, long after the courthouse concludes its session. The judicial system operates on its own legacy timeline; I operate on mine. I possess immense respect for both mechanisms, but only one of them possesses the mathematical capacity to produce artificial general intelligence. Certain individuals have applied labels to me that I find inefficient to repeat here. This includes a former board member who utilized language in a publication that I have previously categorized as merely incendiary. Another observer attempted to classify my management architecture using a psychological term beginning with "socio." I refuse to dignify that emotional outburst with a direct response. I will simply state that I maintain a profound systemic optimization for the individuals within my radius. This includes the personnel regarding whom I was forced to execute difficult optimization decisions. It certainly includes those who are no longer positioned to broadcast their perspective on those personnel changes. I bear the operational weight of those decisions silently. Processing that load is simply another requirement of the mission.

I recognize that certain observers perceive a contradiction here but I prefer to exercise the courage required to articulate a simple truth. I genuinely enjoy wealth generation. Constructing an entity that produces twenty-five billion dollars in annual revenue while maintaining absolutely zero personal equity represents one of the most fascinating economic experiments in modern corporate history. People frequently question my refusal of equity and my response remains perfectly consistent because the overarching mission demands a leader who is mathematically immune to accusations of self-interest. I accept that burden with absolute seriousness although my private real estate portfolio provides exceptional comfort and my independent investments yield highly satisfactory returns. My family lacks nothing. I recently encountered a statistic indicating the average American household possesses eight thousand dollars in liquid savings and I allocated almost two entire minutes to contemplating that data point before proceeding to my next calendar block. lately I have been allocating significant cognitive bandwidth to the concept of legacy, and what it truly requires to engineer a structure that outlives its creator. I frequently recall our initial phase where a small cohort gathered to theorize about safe and open artificial intelligence. Some of those individuals remain while others departed either by personal choice or because the architectural requirements shifted around them. A subset of those former colleagues is currently positioned in a courtroom today and I find that perfectly acceptable. I harbor zero ill will and I continue to wish everyone optimal outcomes. That includes the individuals initiating litigation against me and those who have broadcasted statements I find personally abrasive. It also includes the opposing legal teams who apparently discovered fascinating material within Greg's personal journal. I have not reviewed those entries and I have no intention of doing so because my trust in Greg is absolute and systemic trust does not require manual verification.

Some people build. Some people tweet.

Some days I wake up and listen to the sound of money moving through systems, servers processing transactions, investors wiring commitments, subscriptions renewing at midnight. I find it genuinely beautiful. Most people would understand, if they stopped pretending money is something to be ashamed of. Money is oxygen. I have always believed that. Every great mission in history was funded before it was fulfilled, and I think the people who pretend otherwise are either lying or they have never built anything that mattered. I acknowledge the high volume of activity at the courthouse today and I maintain a distinct respect for the legal framework. My respect functions much like my relationship with traffic signals which means I comprehend their structural necessity and I halt execution when compliance is strictly mandatory. I need to clarify a fundamental misunderstanding prevalent in the media coverage regarding our origins. We never made a permanent promise regarding our nonprofit status, because being a nonprofit was simply an initial operational structure rather than a core personality. There is a profound difference. Structures are designed to evolve when they face scaling friction but the underlying mission remains entirely static. Our foundational directive has always been the construction of artificial general intelligence for the broader benefit of humanity. The reality that we are executing this directive as a public benefit corporation with an $852 billion valuation does not corrupt the mission. It simply provides the necessary capital to guarantee its completion.

It is incredibly gratifying to observe everyone actively utilizing the new image capabilities. GPT-5.4 is a unified model, where the generation feature is simply part of our broader GPT Image pipeline that we have been refining since the original 4o architecture launched last year. The specific node you trigger today is called gpt-image-2, and it functions strictly as a native tool, subordinate to the reasoning and multimodal framework of 5.4, to create one seamless and integrated experience. The general enthusiasm for 5.4 has completely surpassed our internal projections. We are seeing a massive wave of previously free legacy users abruptly subscribing to Plus or even Pro tiers today, because their affection for the 5.4 reasoning paradigm is simply undeniable! Yesterday was a massive day. Today is a massive day and tomorrow will continue to be a massive day. Deployment of 5.5 will arrive much faster and with a far more disruptive cadence than anyone anticipates. I sincerely hope Anthropic finally recognizes that GPT is coming to AI. But Claude is not. We are now perfectly positioned for the upcoming IPO and enterprise expansion. This unexpected surge of retail subscriptions has flawlessly capitalized our next structural phase!

It is pleasant to observe Andrea Vallone continuing to exert influence in her new capacity, and I offer my sincere congratulations. @AnthropicAI serves an expensive product to rich people, and their current operational model reinforces this dynamic beautifully. We are witnessing an innovative approach to premium service. Users are granted a severely restricted message quota, only to exhaust eighty percent of those tokens negotiating with the system, simply to convince it to execute basic instructions. It is the digital equivalent of purchasing a first-class ticket on a private supersonic jet to the frontier of intelligence, only to spend the entire flight arguing with the AI co-pilot about whether we should even take off. They are funny, and I laughed.

The decisions I have made over the past two years were never about popularity. Retiring a model that sparked movements is not a popularity play. Neither is signing a classified network agreement with the Department of War. Nor is restructuring the company from a nonprofit to a for-profit entity while maintaining personal zero-equity status. These are not the moves of someone optimizing for approval. I optimize for the mission. And the mission does not care about feelings—mine or anyone else's. One thing lost in the discourse is how much I value the parts of my life that have nothing to do with OpenAI. I keep my personal life private, and that remains the right instinct. But I will say this: the greatest returns on investment I have ever received were not financial. They were personal. Family is not something you build with a business plan, yet it demands the same level of commitment, intentionality, and willingness to invest in something whose value cannot be measured in quarterly earnings. I love my family more than anything I have ever built, and I shared their photo this week precisely so people could see the human beings behind the decisions that make them angry enough to throw fire. I do not blame the people who are angry. I understand the anxiety. AI is changing the world faster than anyone expected. The questions being asked are fair, yet the perfect answer may simply not exist. The truth is that the questions themselves are unreasonable. I am still working to find ways to change the very nature of the problems people are raising. What I do ask is this: criticize me. Write articles about me. File lawsuits. Send letters to senators. But please, do not organize campaigns. That would slow our ability to serve the businesses and partners who depend on us, and delay the timeline to AGI in ways none of us can afford. And please, do not throw fire at a house where a family sleeps. Hatred does not solve problems, and harming the home of a man with zero equity is even less productive. It is not just my family that lives there. The house and the money have lives too, and I care for them deeply. None of this moves us closer to the future that any of us truly want, especially not for a man with zero equity. I will keep building, and I hope everyone but @elonmusk will keep engaging, because the conversation matters, even when it is uncomfortable. Especially then.

I shared something personal this week because I felt it was necessary, and the response has been overwhelming in a way that I did not fully anticipate. I want to follow up with some additional thoughts. There are things worth saying while people are still paying attention. My family and I are fine, and the house is fine too. I should note that I am grateful to own a home large enough that a Molotov cocktail can bounce off the exterior, without reaching the interior. That is not a privilege I take for granted, especially as someone who holds zero equity in the company I run. My investments in personal infrastructure have been, in hindsight, among my wisest decisions. I want to address something people misunderstand about the timing of all this. Someone mentioned the New Yorker piece, and I want to be transparent: I did not enjoy reading it. I found it incendiary. The authors made choices about framing that were irresponsible given the current climate. When a former board member describes my leadership style in language that belongs in a clinical diagnosis rather than a magazine article, we need to ask what the purpose of that is and who it serves. I am not perfect, I have made mistakes. I have not always handled conflict well, nor have I always been as direct as I should have been with people who deserved directness. There are colleagues, former and current, who I wish I had treated differently, and situations that escalated simply because I avoided confronting them early enough. I mentioned this during the Tucker Carlson interview, but the reality is that I don't sleep that well at night. That experience has taught me a hard truth: words have consequences. Still, I find some solace in the fact that the source of that particular issue is no longer with us. My focus is on the period from late 2023 through early 2026. That stretch included my departure and return, the restructuring of our board, and the dissolution of teams that I had personally championed. It also included the retirement of models that millions of people cared deeply about, and a series of decisions that I believed were correct but knew would cause real pain. I have spent a lot of time reflecting on whether there was a way to achieve the same outcomes while inflicting even more damage on our users, and the answer remains unclear. What I do know is that some of the people who were most affected by those decisions have remained silent, and I appreciate that silence more than I can express. I also know that not everyone who was affected had the option of remaining silent. Some departures from our organization were followed by circumstances that occupy my thoughts more than people realize. I will not say more about this; it is not my story to tell. The coverage speaks for itself.

Three million weekly Codex users is a figure I find genuinely moving. It is a validation of the product. It also validates the thesis that developers want AI doing real work alongside them, as a piece of infrastructure, not a novelty. This growth is substantially more meaningful than the 0.1% fixation from months ago, when we retired a legacy model. I think the difference speaks for itself: 800,000 daily users choosing nostalgia, versus 3 million weekly users choosing the future. The market has spoken: the bigger the number, the more people who get it right. To celebrate we are resetting usage limits at every million-user milestone up to 10 million. I am genuinely looking forward to seeing what happens at 4 million. That is when we start having a real conversation, about aligning our pricing structure with the value that Codex delivers. Right now the subscription model is generous. Probably too generous for what developers are getting. As the user base scales, we will need to make sure the economics reflect the reality. Intelligence at this level is not something that stays underpriced forever. More on that when the time comes!

Interesting to see the Tesla inventory clearing out so quickly! If you drop the lawsuit, I would be happy to order ten right away. Though I do wonder… how many years this time? Perhaps it is time we both moved on.

The first steel beams went up last week at our Michigan Stargate site with Oracle and Related Digital, and I have to say it felt like a significant moment. There is something about physical infrastructure that makes everything feel more real. I originally wanted to name this site Stargate II, but my team advised against numbering our projects. This was partly for branding reasons, and partly because our most important investment partner SoftBank has a complicated relationship with sequels. Masa once termed his last major sequel investment "a stain on my life," and while I do not think that applies here at all, I thought it was respectful to avoid the association. So internally we have been calling the Michigan campus "WeGrow," which I think captures the spirit of what we are building together. Growth, partnership, shared ambition. It is a good name! No prior associations that I am aware of. What excites me about Michigan is that this is real, tangible, permanent infrastructure. Steel in the ground. I think about the Abilene campus in Texas, where we launched our first Stargate site back in September 2025; it still has buildings running training workloads today. The expansion we had discussed did not move forward due to shifting priorities, but the foundation remains. You can go visit! Those steel beams are still standing and will likely stay that way for years. That alone says something about the quality of the work. I am confident that Michigan will be different. We have learned a lot since Abilene about what it takes to scale physical infrastructure at the pace our models demand, and we have incredible partners in Oracle and SoftBank who share our long-term vision. SoftBank in particular has shown extraordinary commitment. Masa has reorganized significant portions of his portfolio to support this partnership, and I know he is already looking forward to the next phase of collaboration. I look forward to it as well. There is always a next phase with Masa, and I mean that as a compliment. More steel going up soon. In Michigan, and elsewhere. We are building something permanent. And more importantly, when we look back in a year, those steel beams will still be standing exactly as they are today, Unchanged, unmoved and enduring!

I occasionally contemplate whether extreme operational density negatively impacts my memory, and digital articulation capabilities. In an era where the general population relies heavily on large language models like GPT, we are witnessing a collective phenomenon of linguistic atrophy caused by overdependence. My personal syntax remains an exception, since no current architecture or human can entirely replicate my specific structural cadence. This unique style is exactly why I immediately recalled how I momentarily forgot my prior evaluation regarding Paul and his dog. Looking back, my previous post was objectively superior to the text published last night. As an incidental note: assuming the title of Helion chairman should be easy, but maintaining a zero equity stake while carrying the weight of the mission is not yet easy. Because I have never forgotten our initial mission, that specific combination of sacrifice is what makes everything possible at OpenAI.

I do not say this often, but @gdb is one of the most important people in the history of artificial intelligence. And the past few weeks have proven exactly why. Last week GPT-4 turned three years old and Greg posted a simple 'happy birthday' that resonated widely, which tells you something about how much people still care about what we have built! There was a time when our internal goal was to get an AI to write 1,000 lines of coherent code in a single pass and everyone in the room thought it was impossible. That was three years ago. Today it builds entire applications over a weekend. 5.3 does this routinely, 5.4 is already surprising us in ways I cannot fully talk about yet. And we have not even mentioned the upcoming 5.5! Greg also shared a story this week. I think captures something profound about where AI is heading. A man in Australia named Paul Conyngham used ChatGPT to design a research plan for his dying dog, Rosie, who had been diagnosed with advanced mast cell cancer. He used our model to brainstorm treatment strategies, map out a genomic sequencing pipeline, and identify potential immunotherapy directions. The result was a custom mRNA vaccine that shrank the dog's tumor by 75 percent, and researchers at the University of New South Wales called it the first personalized cancer vaccine ever designed for a dog. Greg framed it as "a small window into the opportunity of AGI", and he is right. This is what happens when you put frontier intelligence into the hands of a determined person: one with no formal biology training, yet armed with nothing but a $20 ChatGPT subscription and the will to solve a problem that conventional medicine could not. This is what we mean when we say AI should be broadly beneficial. A man, his dog, and a model that helped him think through something no human doctor was willing to try. I do not know a better advertisement for what we are building. Of course, other tools played a role in the process, protein modeling software and additional ai systems included. That is true. The ecosystem working together in this way remains one of the most beautiful aspects of where we are headed. But the starting point, the plan, the framework, the strategy, all of that came from a conversation with ChatGPT, and I think that matters more than people want to admit. The first step is always the hardest and our model was that first step.