Megalomania? Hypergrowth on credit: OpenAI's (ChatGPT) 100 billion gamble against economic history

The dissonance between technological promise and economic reality

OpenAI has set out to redefine the boundaries of artificial intelligence. But while the company is making ambitious predictions about the capabilities of its models, it is simultaneously planning revenue growth that shatters all historical precedents. A recent analysis by Epoch AI paints a remarkable picture: OpenAI aims to increase its revenue from $13 billion in 2025 to $100 billion by 2028. This equates to a required annual growth rate of 97 percent over three years. By comparison, even the fastest-growing companies in recent tech history, such as Tesla and Meta, needed seven years to make the leap from $10 billion to $100 billion in annual revenue, and Google even a full decade. OpenAI wants to reach this milestone in just three years, a pace for which, according to Epoch AI, there is no historical precedent.

These ambitions raise fundamental questions. Is this the justified extrapolation of a technological revolution whose transformative potential rewrites the rules of the market economy? Or are we witnessing a repetition of historical patterns where inflated growth expectations and massive infrastructure investments inevitably lead to overcapacity and economic disruption? The answer likely lies somewhere in between and requires a nuanced examination of the technological, economic, and structural factors that determine OpenAI's growth trajectory.

This article analyzes OpenAI's growth strategy within the context of economic history, examines the underlying market mechanisms, and assesses the likelihood of the company achieving its objectives. It highlights both the innovative strengths and the structural risks associated with such an aggressive expansion strategy. The analysis is divided into eight sections: a presentation of the historical development, an identification of the key drivers of the current AI boom, an assessment of the current situation, comparative case studies, a critical evaluation of the risks, an outlook on potential development paths, and concluding strategic implications.

Related to this:

• Profit before principles? The sex turn – ChatGPT gets dirty and why OpenAI is now focusing on erotica

From research lab to the world's most valuable startup

OpenAI's story is inextricably linked to the rise of large language models and the wider commercialization of artificial intelligence. Founded in 2015 as a non-profit research organization, the company initially positioned itself as a counterweight to the major technology corporations, pursuing the goal of developing artificial general intelligence for the benefit of all humankind. The founders, including Sam Altman and Elon Musk, recognized early on that developing advanced AI systems would require enormous capital resources.

The decisive turning point came in 2019 with the transformation into a hybrid structure, combining for-profit and non-profit elements. This realignment enabled OpenAI to secure its first investment of one billion dollars from Microsoft. The partnership with the software giant proved strategically valuable: OpenAI gained access to Microsoft's Azure cloud infrastructure and the necessary computing resources, while Microsoft, in return, received exclusive access to OpenAI's technology.

The company's revenue initially grew moderately in the following years. In 2020, OpenAI generated only $3.5 million in revenue, but a year later, this figure had risen to $28 million. The breakthrough came in November 2022 with the release of ChatGPT, a chatbot based on GPT-3.5, which reached one million users within five days and surpassed the 100 million user mark within two months. This viral success transformed OpenAI overnight from a research lab into a commercial powerhouse.

Revenue growth accelerated dramatically. In 2023, OpenAI surpassed the one billion dollar mark in annual revenue for the first time, reaching 1.6 billion dollars. In 2024, revenue more than doubled to 3.7 billion dollars. For 2025, the company projects annualized revenue of 13 billion dollars, representing a 251 percent increase over the previous year. This momentum is driven by a growth rate of approximately 3.2 times per year since the end of 2023.

Parallel to revenue growth, the company's valuation soared to dizzying heights. A funding round in March 2025 valued OpenAI at $300 billion. Just a few months later, in October 2025, a secondary stock sale to investors such as SoftBank, Thrive Capital, and T. Rowe Price pushed the valuation to $500 billion. This made OpenAI the most valuable startup in the world, surpassing even Elon Musk's SpaceX.

This historical development illustrates the extraordinary speed with which OpenAI has evolved from a research project to one of the dominant players in the global AI industry. At the same time, it raises the question of whether these valuations are based on realistic assumptions about future growth and profitability, or whether they represent an overvaluation reminiscent of previous technology bubbles.

Drivers, players and the mechanics of the AI ​​market

The current AI boom is driven by a complex interplay of various factors. At its core lies technological innovation itself: Large language models have made remarkable progress in recent years in natural language processing, logical reasoning, and solving complex tasks. These capabilities open up application possibilities in virtually every economic sector, from automating customer service and software development to scientific research.

The key players can be divided into several categories. First and foremost are the developers of large language models such as OpenAI, Google with Gemini, and Anthropic with Claude. These companies compete for technological leadership and market share, with OpenAI currently holding a dominant position with ChatGPT. ChatGPT's market share in the field of AI assistants is estimated at 62.5 percent.

A second key group consists of infrastructure providers. Nvidia dominates the AI ​​accelerator market with a market share of approximately 95 percent. The company's graphics processors, particularly the H100 and A100 series, have become indispensable for training and running large language models. Nvidia is profiting massively from the AI ​​boom and has multiplied its valuation in recent years. However, AMD and Broadcom have recently entered the market, attempting to challenge Nvidia's dominance.

Cloud providers such as Microsoft Azure, Amazon Web Services, and Oracle form a third important category of players. They provide the computing power required for training and operating AI models. OpenAI's close partnerships with Microsoft and Oracle are of particular significance in this context.

The economic incentives driving these players are multifaceted. For OpenAI and its competitors, it's about establishing a dominant market position in a technology segment with the potential to transform large parts of knowledge work. McKinsey estimates that generative AI could contribute between $2.6 and $4.4 trillion annually to global economic output. Given such forecasts, even investments in the hundreds of billions seem justified.

For infrastructure providers like Nvidia, this creates a direct demand for their products. The market mechanism follows a self-reinforcing logic: the more capital flows into the development of larger and more powerful models, the greater the demand for computing power and thus for chips. This dynamic has led to a veritable arms race, in which companies like OpenAI are concluding long-term supply contracts worth hundreds of billions of dollars.

Another key driver is the availability of capital. The low interest rates of recent years and the general euphoria surrounding artificial intelligence have led investors to pour enormous sums into AI startups. OpenAI alone closed a $40 billion funding round in the first half of 2025 and secured an additional $4 billion revolving credit facility. This influx of capital allows the company to pursue its ambitious expansion plans despite massive operating losses.

Regulatory frameworks also play a role, albeit an ambivalent one. On the one hand, there are efforts in key markets like the European Union to more strictly regulate AI systems, which could increase development costs. On the other hand, governments, particularly in the US, actively support AI development. The Stargate project, the largest AI infrastructure initiative in history with a total budget of $500 billion over four years, was launched with strong support from the Trump administration.

The underlying market mechanisms exhibit characteristics typical of technology markets. It is a market with high fixed costs and low marginal costs: developing a large language model costs hundreds of millions to several billion dollars, while the cost of answering a single user query is comparatively low. This leads to strong economies of scale and favors the emergence of oligopolies or even monopolies.

At the same time, this is a market with network effects: the more users a platform like ChatGPT has, the more valuable it becomes due to the generated data and user feedback, which can contribute to improving the models. However, these network effects are less pronounced in the case of large language models than, for example, in social networks, since users can relatively easily switch between different providers if a competitor offers a better model.

Indicators of unprecedented expansion and its limits

OpenAI's current situation is characterized by a discrepancy between impressive growth and massive financial losses. In the first half of 2025, the company generated $4.3 billion in revenue, already 16 percent higher than its total revenue for the previous year. At the same time, however, OpenAI recorded an operating loss of $7.8 billion. This loss margin amounts to 181 percent of revenue, illustrating that for every dollar earned, the company is spending nearly two dollars more.

The main cost drivers are clearly identifiable. Research and development alone consumed approximately $6.7 billion in the first half of 2025. A significant portion of this is attributable to the computational costs for training new models and operating ChatGPT. Estimates for the training costs of the next model generation vary considerably: while GPT-4 was estimated to cost between $100 million and $200 million, the training costs for GPT-5 could range from $500 million to $2 billion, depending on the source. These exponentially increasing development costs represent a key challenge.

Added to this are personnel costs, which are also rising rapidly. OpenAI granted its employees stock options worth $2.5 billion in the first half of 2025, almost twice as much as in the entire previous year. The intense competition for AI talent is driving up salaries and forcing companies to offer generous compensation packages.

ChatGPT's user base continues to grow dynamically. In October 2025, the platform recorded between 700 and 800 million weekly active users. This represents a doubling compared to February 2025, when the number stood at 400 million. The platform processes 2.5 billion queries daily and ranks fifth among the most visited websites worldwide.

The central problem, however, lies in the conversion rate. Only five percent of users pay for a subscription, be it ChatGPT Plus for $20 per month or ChatGPT Pro for $200 per month. This equates to approximately 40 million paying users. Even this comparatively low conversion rate is above the average for the generative AI industry, where only three percent of users are willing to pay. Nevertheless, the fact remains that 95 percent of the user base is currently not generating any direct revenue.

Approximately 75 percent of total revenue comes from consumer products, primarily ChatGPT subscriptions. While the enterprise customer business is growing, it remains relatively small. In June 2025, OpenAI reported three million paying business customers for its ChatGPT Enterprise, ChatGPT Team, and ChatGPT Edu products. By September, this number had risen to five million. Although this represents healthy growth, the B2B segment still lags significantly behind the consumer business.

A valuation of $500 billion implies a price-to-sales ratio of approximately 38.5 based on projected revenue of $13 billion for 2025. For comparison, software companies are typically valued at two to four times their annual revenue. Even high-quality, high-growth SaaS companies rarely achieve multiples above ten. OpenAI's valuation is therefore many times higher than historical averages and reflects investors' extreme growth expectations.

These expectations are based on the assumption that OpenAI can reach its revenue target of $100 billion by 2028. To achieve this, the company would have to overcome several challenges: The number of paying users would have to increase dramatically, possibly to 200 to 300 million. At the same time, new revenue streams would have to be developed, such as advertising, e-commerce integrations, or high-priced productivity tools for businesses.

The infrastructure commitments OpenAI has made intensify the pressure to succeed. Contracts with Nvidia, AMD, and Broadcom total approximately $1.3 trillion over a decade. The Stargate project envisions investments of $500 billion over four years. These commitments far exceed current and even projected revenues and necessitate continuous capital injections from investors or a significantly faster improvement in profitability.

Here you will learn how your company can implement customized AI solutions quickly, securely and without high entry barriers.

A managed AI platform is your all-inclusive, worry-free solution for artificial intelligence. Instead of dealing with complex technology, expensive infrastructure, and lengthy development processes, you receive a ready-made solution tailored to your needs from a specialized partner – often within just a few days.

The key advantages at a glance:

⚡ Rapid implementation: From idea to ready-to-use application in days, not months. We deliver practical solutions that create immediate added value.

🔒 Maximum data security: Your sensitive data stays with you. We guarantee secure and compliant processing without sharing data with third parties.

💸 No financial risk: You only pay for results. High upfront investments in hardware, software, or personnel are completely eliminated.

🎯 Focus on your core business: Concentrate on what you do best. We take care of the entire technical implementation, operation, and maintenance of your AI solution.

📈 Future-proof & scalable: Your AI grows with you. We ensure continuous optimization and scalability, and flexibly adapt the models to new requirements.

More information here:

• The Managed AI Solution - Industrial AI Services: The Key to Competitiveness in the Services, Industry and Mechanical Engineering Sectors

Lessons from the rise of digital giants and their limitations

A look at comparable companies and their growth trajectories offers valuable insights into the feasibility of OpenAI's ambitions. Google, now Alphabet, reached the $100 billion mark in annual revenue within ten years of its IPO in 2004. The company benefited from near-monopolistic access to the lucrative search engine market and was able to achieve high margins through advertising revenue. Google's business model was based on low marginal costs and strong network effects, which enabled continuous profitability.

Meta, formerly Facebook, also needed seven years to jump from $10 billion to $100 billion. Meta, too, benefited from strong network effects and an advertising-based business model with high margins. Crucial to Meta's success was its ability to monetize a massive user base, initially on desktop and later on mobile devices. The acquisition of Instagram and WhatsApp further expanded its user portfolio.

Tesla presents an interesting case study, as the company operates in a capital-intensive industry with lower margins. Tesla also reached its $100 billion revenue target in approximately seven years, but benefited from a period of exceptionally high valuations for electric vehicle manufacturers and a charismatic CEO who embodied the brand. Tesla struggled with profitability issues and negative cash flow for years before finally crossing the break-even point.

A comparison with these companies reveals both parallels and important differences to OpenAI. All three companies benefited from technological innovations that transformed existing markets. All three had strong brands and charismatic leaders. However, Google and Meta achieved profitability significantly earlier in their development than OpenAI. Tesla, on the other hand, recorded losses for extended periods but was able to bridge these gaps through continuous capital raising.

A critical difference lies in the nature of economies of scale. At Google and Meta, the cost per user decreases significantly as the user base grows because infrastructure costs remain relatively constant. At OpenAI, however, computing costs rise almost proportionally with usage, as every request to ChatGPT consumes computing resources. CEO Sam Altman admitted that OpenAI is losing money on its $200 ChatGPT Pro subscription because users are using the service more intensively than anticipated. This points to a fundamental problem: Without dramatic cost reductions, growth does not automatically translate into improved profitability.

Another relevant comparison concerns companies that failed in their attempts to sustain extremely rapid growth. During the dot-com bubble of the late 1990s, hundreds of internet companies emerged with similarly ambitious growth forecasts. The majority failed because revenues did not keep pace with expectations and investors eventually lost patience. The telecommunications sector also experienced massive misinvestments when companies built fiber optic networks with a capacity that far exceeded actual demand.

Chinese AI development offers another interesting point of comparison. DeepSeek, a relatively unknown Chinese startup, caused a stir in early 2025 when it released a language model that could compete with leading Western models, but reportedly cost only a fraction of the development costs. DeepSeek's R1 model is said to have cost just $5.6 million, compared to over $100 million for GPT-4. If it turns out that comparable performance can be achieved with significantly fewer resources, this challenges the assumption that massive investments in computing power are the only path to advanced AI systems.

Related to this:

• Is ChatGPT from OpenAI and Google Gemini AIaaS – Artificial Intelligence as a Service?

Fractures, uncertainties, and the anatomy of possible maldevelopments

The risks associated with OpenAI's growth strategy can be divided into several categories. First, there are significant technological uncertainties. The so-called scaling laws, according to which larger models with more training data and computing power automatically improve, may be reaching their limits. There are indications that newer models no longer show the same performance leaps as previous generations. For example, OpenAI's GPT-5 reportedly consumed less computing power during training than GPT-4.5, without delivering significantly better results. This could indicate that the simple scaling laws are losing their validity and that new approaches are needed.

The competitive landscape is intensifying. Google with Gemini, Anthropic with Claude, and Meta with its Llama models are all investing heavily in the development of competing systems. Each of these players possesses significant resources and established distribution channels. Google can integrate Gemini into its search and productivity tools, while Meta can integrate its models into Facebook, Instagram, and WhatsApp. OpenAI lacks a comparable ecosystem, which increases its reliance on ChatGPT as its primary distribution channel.

The cost structure presents a structural problem. The computational costs for running large language models are enormous and increase with usage. OpenAI is estimated to spend 60 to 80 percent of its revenue on computation costs alone. This leaves little room for profitability, especially considering the additional costs for personnel, research, and operations. A significant reduction in inference costs would be necessary, but whether and when this will be achieved remains uncertain.

The reliance on a few infrastructure providers carries additional risks. Nvidia almost completely controls the market for AI accelerators, giving the company considerable pricing power. While OpenAI is attempting to reduce this dependence through contracts with AMD and Broadcom, these alternatives require time to build up production capacity. Should chip supply bottlenecks or drastic price increases occur, this could significantly impact OpenAI's expansion plans.

Regulatory risks are increasing. Questions regarding copyright on training data, data protection, and liability for AI-generated content remain largely unresolved. Should courts or legislators decide that AI companies must pay for the use of copyrighted training data, this could dramatically alter the cost structure. Stricter data protection regulations or restrictions on certain use cases could also stifle growth.

The risk of an infrastructure bubble is real. The historical parallels to the telecommunications bubble of the late 1990s are striking. Back then, massive capital inflows led to the construction of network capacity that far exceeded actual demand. When the bubble burst, 85 to 95 percent of the laid fiber optic cables remained unused, and dozens of companies went bankrupt. Today, observers see similar patterns in the data center boom: massive capacities are being built, but their full utilization is uncertain. Should demand for AI services fall short of expectations, many of these investments could become worthless.

A valuation of $500 billion implies extremely optimistic assumptions. Investors buying in at this valuation apparently expect an IPO at a valuation of over $1 trillion within two to three years. This would make OpenAI one of the ten most valuable publicly traded companies in the world. By comparison, Apple took decades to reach such a valuation and has massive cash flows and an established product portfolio. OpenAI, on the other hand, is incurring heavy losses and is dependent on a single product.

The social and environmental costs of AI expansion are increasingly being discussed. The energy consumption of large language models is considerable. The Stargate project, for example, is projected to require ten gigawatts of electricity, equivalent to the energy needs of approximately 7.5 million households. In the context of the climate crisis, this raises questions about the sustainability of such investments. Furthermore, negative social impacts, such as the automation of jobs, could lead to political opposition.

Scenarios between breakthrough, stagnation and correction

The future development of OpenAI and the broader AI industry can be outlined along several scenarios. In the optimistic scenario, OpenAI succeeds in achieving its ambitious growth targets. This would require several conditions to be met: Technological development continues, with new model generations offering substantial improvements. The conversion rate of paying users increases significantly, potentially to 15 to 20 percent, which would translate to 120 to 160 million paying subscribers. New revenue streams, such as advertising, e-commerce, and high-priced enterprise products, are successfully developed and contribute significantly to overall revenue. Inference costs decrease considerably due to technological advancements and increased competition in the chip market. In this scenario, OpenAI would become profitable and could go public at a valuation exceeding one trillion dollars.

In the moderate scenario, OpenAI continues to grow but falls short of its most ambitious goals. Revenue might reach $40 to $60 billion by 2028 instead of $100 billion, which would still represent exceptional growth. However, profitability remains difficult to achieve as costs keep pace with growth. OpenAI would need to rethink its infrastructure plans and potentially renegotiate some contracts. Its valuation would be adjusted, possibly to $200 to $300 billion. An IPO would still be possible, but at more modest valuations. In this scenario, the AI ​​market establishes itself as an oligopoly with several large players competing for market share.

In the pessimistic scenario, OpenAI faces significant growth obstacles. Technological development slows, and new models fail to offer sufficient added value compared to existing solutions. Competitors like Google and Anthropic gain market share. The conversion rate stagnates at low single-digit percentages. At the same time, costs remain high or even continue to rise. In this scenario, OpenAI could struggle to secure further funding rounds at attractive valuations. The company would have to drastically reduce its spending and potentially sell assets. Its extensive infrastructure commitments would become an existential burden. This scenario could trigger a broader correction across the entire AI sector, similar to the bursting of the dot-com bubble.

A disruptive scenario would be the commercialization of fundamentally more efficient AI architectures. Should approaches like the techniques demonstrated by DeepSeek gain wider application, this could fundamentally alter the industry's cost structure. In this case, the massive investments in traditional scaling would lose value. OpenAI would have to adapt its strategy and could lose its lead in the process. At the same time, this would accelerate the democratization of AI and allow more competitors to enter the market.

Another important element is the development of AI agents capable of autonomously executing complex tasks. If reliable agents can be developed to function as virtual employees and enable companies to achieve significant productivity gains, this could usher in a new phase of growth. OpenAI is positioning itself for this market, but the technological challenges are considerable. Current AI systems are prone to hallucinations and errors, which limits their reliability for critical business processes.

Regulatory developments will also play a key role. Governments in the US, Europe, and China are developing different approaches to AI regulation. Stricter regulations could stifle innovation but also foster greater trust and broader acceptance. Conversely, a regulatory vacuum could lead to abuse and societal disruption, ultimately prompting harsher interventions.

The geopolitical dimension is gaining importance. The AI ​​competition between the US and China is increasingly perceived as a strategic confrontation. Export controls, investment restrictions, and government support programs could significantly influence the competitive dynamics. The Stargate project is explicitly designed as a contribution to American technological leadership.

Between visionary ambition and economic disillusionment

OpenAI's plan to increase revenue from $13 billion to $100 billion within three years represents one of the most ambitious growth plans in the history of the technology industry. Analysis shows that while this plan is not impossible, it would require a multitude of favorable conditions, the simultaneous occurrence of which must be considered unlikely.

OpenAI's strengths are undeniable. The company boasts technological leadership in large language models, a strong brand, and a massive user base. ChatGPT has become synonymous with generative AI, much like Google is synonymous with internet search. Partnerships with Microsoft and Oracle ensure access to essential infrastructure resources. Its capital base has been strengthened through several funding rounds.

At the same time, the challenges are immense. The low conversion rate of paying users, the high and ever-increasing development costs, the intensified competition, and the structural profitability problems pose significant obstacles. The infrastructure commitments undertaken far exceed the foreseeable revenues and create enormous pressure to succeed.

Several implications arise for policymakers. First, the massive government support for AI infrastructure should be critically examined. The Stargate project may be symbolically valuable, but its economic viability is questionable when private investors risk hundreds of billions without a sound business case. Second, regulatory frameworks should be developed that enable innovation while simultaneously addressing risks. Third, the energy issue must be resolved: The massive electricity demand of AI data centers clashes with climate goals and requires coordinated solutions.

For business leaders, this development means that AI investments should be approached strategically, but without unrealistic expectations. The productivity gains from AI are real, but they will materialize gradually and require significant organizational adjustments. Companies should experiment, but not base their business model on immature technologies.

Investors are faced with the question of appropriate valuation. The current valuation of $500 billion only seems justified if OpenAI not only meets but exceeds its growth targets and simultaneously achieves profitability. The risk-reward ratio is extremely unfavorable for late investors. Early investors, who entered the market at significantly lower valuations, can realize substantial profits even with moderate success.

The long-term significance of OpenAI and broader AI development for the global economy should not be underestimated, regardless of whether the company achieves its specific revenue targets. Large language models will transform parts of knowledge work and enable significant productivity gains. The question is not whether this transformation will occur, but how quickly and which companies will benefit from it.

History teaches us that technological revolutions are often accompanied by financial excesses. The railroad, electricity, automobile, and internet revolutions all saw phases of massive overinvestment followed by painful corrections. Yet these technologies ultimately proved transformative. The investors who profited the most were often not those who built the infrastructure, but those who used that infrastructure to develop innovative business models.

OpenAI is at a turning point. The company must prove that it can not only develop impressive technology but also translate it into a profitable business model. The next two to three years will be crucial. Should OpenAI fail to meet its goals, the repercussions will extend far beyond the company and shake the entire AI sector. Conversely, if it succeeds, it would rewrite the rules of corporate growth and potentially mark the beginning of a new era in business history.

The key finding of this analysis is that OpenAI needs new scaling principles, not only for the performance of its AI models, but above all for its own business model. The laws of physics and mathematics that govern the training of neural networks are one challenge. The laws of economics and the market that determine how a company can grow sustainably and become profitable are at least as significant. OpenAI must master both to realize its vision.

☑️ Our business language is English or German

☑️ NEW: Correspondence in your native language!

 

I and my team are happy to be available to you as your personal advisor.

You can contact me by filling out the contact form here wolfenstein@xpert.digital:or simply call me at +49 7348 4088 965. My email address is

I'm looking forward to our joint project.

 

 

☑️ SME support in strategy, consulting, planning and implementation

☑️ Creation or realignment of the digital strategy and digitization

☑️ Expansion and optimization of international sales processes

☑️ Global & Digital B2B trading platforms

☑️ Pioneer Business Development / Marketing / PR / Trade Fairs

Xpert.Digital possesses in-depth knowledge across various industries. This allows us to develop tailored strategies precisely aligned with the requirements and challenges of your specific market segment. By continuously analyzing market trends and monitoring industry developments, we can act proactively and offer innovative solutions. The combination of experience and expertise generates added value and provides our clients with a decisive competitive advantage.

More information here:

• Benefit from Xpert.Digital's 5 areas of expertise in one package – starting from just €500/month