#AI startups

#AI startups

Anthropic, the maker of Claude and last publicly valued at a now-quaint $380 billion, is raising a new round of funding-the company's reportedly looking to rake in as much as $50 billion at a valuation in the $900 billion ballpark. And talking to brokers, investors, and founders about it, they all had the same, clear message: This is not normal.

“Yes, you should. You can go there for a bit and then come back, but you should at least go.” Graham's message landed at a moment when Stockholm's startup ecosystem is attracting increasing attention from investors and founders. According to global data intelligence platform Dealroom, the Swedish capital was home to more than 1,800 startups as of 2025, with a combined enterprise value of $236 billion.

For most companies, there's roughly a 12-month period where the business is at its peak value, and then it crashes out. The companies that capture generational returns are often the ones where someone spies that moment instead of assuming the good times will get even better.

Initially, everybody I asked in the city was certain that this was satire, perhaps the workings of Sacha Baron Cohen or a stunt by union activists; after all, the website also lauds the value created by James Dyson, Roger Federer, and the CEO of Chobani (for having "popularized Greek yogurt"). I was reminded of how several years ago, the faux-conspiracists of the Birds Aren't Real movement rallied outside Twitter's headquarters to critique dangerous social-media rabbit holes.

On January 28th, a 30-person U.S. startup called Arcee AI released Trinity Large, a 400-billion-parameter sparse Mixture of Experts model that challenges Meta's Llama 4 Maverick and Chinese models like Z.ai's GLM-4.5. Trained in just six months for $20 million using 2,048 NVIDIA Blackwell B300 GPUs, Trinity represents a significant achievement in democratizing frontier-grade model development. What sets Trinity apart is its commitment to the Apache 2.0 license-a truly open-source alternative to Meta's proprietary Llama license.

Fast growth is exhilarating. It is also unforgiving. Especially in AI, many companies are seeing hyper-growth, changing the leadership job faster than many founder-CEOs expected. What once required deep personal involvement suddenly demands scale and breadth. The question for leadership is how to adapt without losing the mission, or the magic, that made the company take off in the first place.

Even before he'd graduated from the University of Bath in 2024, Arnau Ayerbe landed a highly coveted role as an AI engineer with JP Morgan - yet he felt limited and uninspired. "I realised very quickly that the person to my right and to my left were going to be me in 20 years, and I didn't want to become that," recalls London-based Ayerbe.