SpaceX, OpenAI and Anthropic: who will win the AI race?

The US IPO market is preparing for what will become one of the biggest technology listing waves in history. Among the names expected to go public in the coming months are three of the most fascinating and – sometimes - controversial companies of the moment: SpaceX, OpenAI and Anthropic.

Excitement is high and understandable: these companies are not simply selling products. They are selling the future. We are talking about artificial intelligence, chat bots, AI agents, autonomous systems, machine reasoning, military applications but also space exploration and orbital infrastructure - themes shaping the next economic era.

Some investors want exposure, others are highly reluctant as behind the extraordinary narratives and eye-watering valuations lies a far more complicated financial reality: the inconvenient truth that today, these companies are also cash-burning machines operating in industries where future profitability remains deeply uncertain.

The real question is no longer whether AI and advanced infrastructure will transform the world. They almost certainly will.

But whether investors buying these IPOs will eventually make money.

So let’s dive in, one at a time!

For years, SpaceX remained one of the mythical private companies that ordinary investors simply could not access.

Under Techo-king Elon Musk, it became the ultimate symbol of Silicon Valley ambition: reusable rockets, Mars colonisation, Starlink satellites and increasingly ambitious visions involving AI infrastructure in orbit.

But despite the rockets and the science-fiction headlines, the real engine of SpaceX today is not space exploration. It is connectivity.

Good news is that Starlink has quietly become the company’s financial backbone, and it is presently generating billions in profits to finance the space business which consumes enormous amounts of capital – as you could imagine – sending rockets to the space cost more than buulding electric cars.

At the same time, SpaceX is evolving into something far broader than a rocket company. The long-term vision increasingly involves:

1. orbital AI infrastructure,
2. space-based computing,
3. satellite-powered connectivity,
4. lunar industry and
5. eventually Martian settlements.

Read that again.

And this is where the story becomes both fascinating and financially terrifying.

Economics of artificial intelligence are becoming extraordinarily expensive. Training large AI models now requires gigantic amounts of electricity, cooling systems, semiconductors and computing capacity.

The AI revolution is colliding with physical constraints: power grids, energy shortages and chip production bottlenecks. Elon Musk’s answer to these earthly limitations is radical. If Earth cannot efficiently provide enough energy and cooling capacity, why not move part of the infrastructure into space, right?

From a technological perspective, the idea sounds visionary.

From a financial perspective, it sounds extremely risky.

And this matters enormously for investors.

Buying SpaceX shares would not simply mean investing in a proven, profitable business. It would mean financing one of the most ambitious and speculative infrastructure projects ever attempted — one that could remain decades away from sustainable profitability.

Adding another layer of complexity, Elon Musk is expected to retain overwhelming voting control, 85% of voting rights, meaning investors would effectively place enormous trust in Musk’s long-term vision and capital allocation decisions.

Elon Musk is the biggest strength and the highest risk for SpaceX.

If SpaceX represents the physical infrastructure of the future, OpenAI represents the cognitive infrastructure.

ChatGPT fundamentally changed public perception of artificial intelligence almost overnight around beginning 2023. Suddenly, AI stopped feeling theoretical and became a mainstream consumer product. It changed the world.

As such, OpenAI now sits at the centre of one of the fastest technology adoption cycles in history.

That is exciting, but behind the excitement lies an extraordinary financial challenge: artificial intelligence is unbelievably expensive to develop and operate. Massive investments in chips, data centres and cloud infrastructure must be made upfront, long before stable profitability can emerge.

Revenues are certainly growing rapidly. But spending is rising just as fast — if not faster – to avoid hitting capacity constraints.

There is no magical recipe: OpenAI will likely lose tens of billions of dollars before potentially reaching sustainable profitability. 

The path toward profitability is filled with uncertainties:

• slowing revenue growth,
• intensifying competition,
• financing pressures and
• rising infrastructure costs.

The competitive landscape is already changing rapidly. OpenAI no longer dominates the market uncontested. Rivals such as Anthropic are increasingly capturing market share, while internal revenue expectations reportedly become more difficult to achieve.

At the same time, OpenAI faces gigantic long-term infrastructure commitments that could eventually reach hundreds of billions — and potentially even exceed $1 trillion — over the coming decade. It is in the middle of a complex AI circle with Big Tech playing both the role of investor and provider of data, chips, computing power.

The numbers are eye-watering: OpenAI is reportedly burning around $10 billion annually - an astonishing number even by Silicon Valley standards – and the competition is coming for its cake.

Anthropic — the company behind Claude — has positioned itself as one of the most credible alternatives to OpenAI.

The company has increasingly focused on AI safety, government applications and enterprise-grade systems, helping it attract developers, institutions and public-sector interest.

Its models are already disrupting established software businesses and contributing to growing anxiety across the software sector.

In many ways, Anthropic represents the next stage of the AI battle: not simply building chatbots, but integrating AI deeply into business operations, coding, defence, cybersecurity and enterprise productivity.

But like OpenAI, there is no magic recipe.

Training increasingly advanced AI systems requires huge investments in semiconductors, cloud infrastructure and data centres. Investments should be made upfront. The product should be ready before selling and the capacity constraints should be managed before hand to avoid any frustration with speed and quality of the service.

And this is where the AI ecosystem starts looking increasingly circular.

Both OpenAI and Anthropic depend heavily on financing and infrastructure support from Big Tech companies such as Microsoft, Amazon and Google. Those same companies then sell cloud services, chips and computing power back to them.

In simple words, the AI ecosystem is increasingly financing itself internally.

And while this structure works during periods of investor enthusiasm and abundant liquidity, the long-term sustainability of the model remains uncertain if revenue growth eventually slows.

Because ultimately, someone must generate real profits and money should flow in from OUTSIDE the circle into it.

That’s the million-dollar question facing investors today.

Everyone agrees that artificial intelligence is revolutionary. But nobody truly knows where the long-term profits will eventually settle.

• Will the biggest winners be the AI model creators like OpenAI and Anthropic?
• Will profits flow primarily toward semiconductor companies like Nvidia?
• Or will cloud giants such as Microsoft, Amazon and Google capture most of the economic value through infrastructure ownership?

At the same time, broader macroeconomic conditions are becoming increasingly difficult to ignore.

• Global yields are rising.
• Governments remain heavily indebted.
• Energy systems are under pressure.
• Geopolitical fragmentation is increasing.
• Financing costs are no longer near zero.

And that matters enormously for highly speculative businesses that may remain unprofitable for years.

The danger is that investors become so captivated by the future narrative that they stop paying attention to valuation.

That is how bubbles form.

Bubble or once-in-a-generation opportunity?

I wish I had the answer.

The truth is that these IPOs could become historic wealth creators, but they could also become symbols of speculative excess if expectations drift too far away from economic reality.

The future probably does belong to artificial intelligence and advanced infrastructure, but even the future can become overpriced.

And when valuations become detached from fundamentals, the eventual pullbacks can be brutal.

The content in this article is provided for educational purposes only. It does not constitute investment advice, financial recommendations, or promotional material.