Welcome to The Nest
Ideas, insights, and conversations shaping AI infrastructure, with Smartbird’s perspective on what matters and what’s next.
IBM and Together AI's $240M Bet Is Really About Inference, Not Training
IBM and Together AI's new $240M inference cluster is a reminder that the infrastructure story has quietly moved from training runs to serving models at scale.
Two Million GPUs Is a Headline. The Real Number Is Whoever Doesn't Have Access to Any.
AWS and NVIDIA are adding 2 million GPUs by 2027–2028. The scale is real news. So is the gap it leaves for everyone not placing an order that size.
The Future Isn't One Cloud
Vendor lock-in doesn't announce itself , it accumulates. Here's why we think portability, not consolidation, is what makes AI infrastructure economics predictable.
Infrastructure Should Remove Friction, Not Add It
The real cost of AI infrastructure usually isn't the GPUs. It's the operational weight teams take on to keep them running. Here's the case for removing it.
Letter To Shareholders
Smartbird has a rare opportunity to build a new company at the start of a major infrastructure transition. Our origin is equally unusual: the company formerly known as Allbirds sold its footwear assets, changed its name, brought in new leadership and raised capital to go all in on AI infrastructure.
Why We're Building Smartbird
Smartbird exists because AI infrastructure forces a bad trade-off between control and simplicity. Here's why we think that trade-off is unnecessary.