We’re partnering with HumanX to offer attendees an extended free trial of CB Insights’ predictive intelligence platform, along with exclusive analysis on the future of the enterprise AI buildout.
Of these, nearly 68% are integrations, pilots, or co-marketing and only 12% are client or vendor relationships, suggesting most companies are still building connectivity and testing use cases.
Three players have emerged as the main hubs for AI partnerships:
Nvidia, embedding its proprietary chip software stack deeply in the AI infrastructure layer
Microsoft and Amazon, extending their positions through cloud distribution, partner programs, and enterprise access
Banks are responding to crypto company momentum by converting traditional deposits into blockchain-based tokens.
Tokenized deposits are digital representations of regular money held at a regulated bank and remain liabilities on the bank’s balance sheet, giving customers the same protections as ordinary deposits.
The tokenized deposit issuance space is the highest-momentum blockchain market by Mosaic, CB Insights’ predictive measure of private company health and success.
Strategic partnerships are powering this movement:
This week, BMO announced a tokenized cash and deposits offering in collaboration with CME Group and Google Cloud
Citi's existing Citi Token Services solution added interbank payments functionality in September
Standard Chartered partnered with Ant International on tokenized deposits in December
Both deals extend Amazon's robotics focus beyond warehouse automation, where it has invested for over a decade since acquiring Kiva Systems in 2012.
CB Insights customers can track who Amazon will acquire next on the company’s interactive Strategy Map.
Not: Hard-coded robots
Physical AI models are making rigid, pre-programmed robots obsolete.
While traditional robots required extensive manual programming to handle specific tasks, physical AI models built on vision-language-action architectures and world models now let robots learn from data and adapt to new environments.
The physical AI model stack spans 4 layers:
Data & simulation: Synthetic data, real-world demonstrations, and virtual training environments
Model approaches: VLMs, VLAs, and world models that give robots perception, action, and prediction
Foundation models: Pre-trained robot intelligence for manipulation, autonomous driving, and multi-robot coordination
Observability: Platforms that monitor deployed robots and feed real-world data back into training
CB Insights' emerging technology insights platform provides all the analysis and data from this newsletter. Our data is the easiest way to discover and respond to emerging tech.
