The AI landscape today is a paradox between an IPO gold rush and pricing reality check. While pricing of AI IPOs are expected to soar above historical norms, the true price of using AI is also rising fast.

OpenAI, Anthropic and SpaceX are planning to go public this year and will probably beat all previous IPO valuations. Anthropic filed to go public Monday and is expected to price its IPO by October. Anthropic's latest $65 billion funding round valued the company at $965 billion.

Two days later, Elon Musk’s SpaceX said it plans to raise $75 billion in its IPO in a few weeks. That would be the largest IPO ever and value the company at $1.75 trillion. While not a pure play AI company, SpaceX acquired xAI this year and Musk is betting big on AI.

OpenAI has not filed for its IPO yet but is widely expected to do so soon.

At the same time, we’ve heard a lot about how expensive AI is becoming, even when it replaces humans. And that’s coming from AI companies and other tech vendors pushing the technology.

“Corporate leaders are starting to question whether soaring AI spending is delivering meaningful returns,” said an Axios story, headlined AI sticker shock hits corporate America.

Axios and others have given examples of that, including:

• Nvidia VP Bryan Catanzaro says that AI compute costs his team more than the employees using it, making AI more expensive than human labor.
• Uber COO Andrew Macdonald called AI costs “harder to justify.” Bloomberg reported that Uber has instituted a monthly $1,500 cap per employee and per agentic coding tool.
• Microsoft’s cancellation of Anthropic Claude Code licenses partly due to pricing concerns.
• An AI consultant told Axios a client recently spent half a billion dollars in a month after failing to put usage limits on Claude licenses for employees.
• Mercor CEO Brendan Foody said the AI startup spends more on AI tokens than employee salaries.
• Light Reading points out how AI demand has caused a spike in the prices of memory chips as Ericsson and Nokia said the technology will make telecom equipment more expensive.

Will any of this slow down AI adoption? That’s unlikely. Recent vendor earnings reports from the likes of Dell, HPE, Nvidia and others show sales increasing despite issues such as supply chain constraints and rising memory costs. And these vendors’ execs credit AI adoption for the increased revenue.

Dell COO Jeff Clarke said Dell booked $24.4 billion in AI orders and HPE CEO Antonio Neri said his company booked $1.8 billion in new AI systems orders last quarter. Gartner predicts AI spending will increase 47% to $2.59 trillion in 2026.

"We don't see a cliff,” Neri said of AI demand.

Bad news for people who are tired of AI talk: we’re still only in the early innings.