In many of the places where we go over the blitzkrieg developments of AI in the last year, we focus a lot on the core advances in tech – what models have “learned” to do, and how – but there’s also a market context that’s useful in revealing the effects on our world.
On that note, I was reading about completed initial public offerings in 2025, to understand more about AI on the business front. What I found was one IPO in the U.S., one in Europe, and one still pending. Each tells its own tale of how companies have harnessed AI to springboard growth in interesting ways.
CoreWeave
The most recent IPO in AI was on our side of the Atlantic, with CoreWeave (CRWV) going public in the spring, with a share price of around $40, issuing 37.5 million Class A shares. The IPO valued the company at around $23 billion.
Since then, the company’s stock spiraled up to around $200 per share before reversing back down to around the $90 range. Ambitious data center plans continue to raise question marks in the investor community.
Another thing to know about CoreWeave is that it has a powerful ally. The firm is in many ways an Nvidia “client state.” It has $100 million in Nvidia funding, where the parent firm holds an estimated 5-7% of CRWV shares. Nvidia also gives CoreWeave special access to its cutting-edge hardware, and provides critical support in scaling operations—such as facilitating GPU access during shortages, and backing CoreWeave’s data center expansion.
LightOn
This IPO technically happened in 2024, and it happened in the Eurozone. LightOn became Europe’s first publicly listed pure-play GenAI company in early November of last year, with a €10.share price, and a market capitalization of about €62 million.
“We are looking forward to deploying the transformative power of our generative AI technology to new customers both in France and internationally and to radically transform business uses.” Said the company’s two CEOs, Igor Carron and Laurent Daudet, in a press statement.
LightOn said it aimed to use available funding to expand sales, innovate its Paradigm platform, and grow its reach internationally. Investors haven’t seen any great stock price increase yet.
Part of the interesting thing about the LightOn IPO is the company’s tech trajectory. The company started out focusing on an OPU (optical processing unit) that was a neat piece of tech. But demand was lackluster, and reportedly, the LightOn people realized that the customer base was more attuned to working on LLMs, so LightOn became an AI model enterprise.
Cerebras
The third IPO? Well, it’s not done yet. However, the company in question is a household name in AI, and we can expect big things to happen there.
Cerebras is the maker of the “dinner chip” GPU, a wafer scale engine (WSE) processor with its own force and gravity in the market. I don’t need to go into detail about what the company does, to those who have a basic understanding of the market landscape. Cerebras is well positioned for whatever the company wants to do next, including going public, where top brass filed for an IPO in September 2024, with the intended Nasdaq ticker CBRS.
That’s a survey of recent and pending IPO action in a sector that has intense interest right now – from investors, and from “regular people” who are just curious about how all of this will shake out as we march toward artificial general intelligence and “the singularity”.
What is this famed event? Well, proponents sometimes describe it as a process where AI gets “smarter” until it surpasses us in capabilities – and humans respond by clinging more tightly to technology, maybe implanting all kinds of wearables, or tethering our biological brains to the cloud. The idea is commonly attributed to Ray Kurzweil, an eminent computer scientist, who is still alive, and still making statements about the impending changes.
“Kurzweil maintains that AGI—AI capable of performing any intellectual task a human can—will arrive by 2029,” writes Al Mamun at Robot Builders Aug. 16. “Despite skepticism, recent advances in large language models, multimodal AI, and autonomous agents support this timeline.”
We’ll see how much of this plays out. Meanwhile, knowing more about the market can help with questions about implementation and the results for business and investment funds.
