The intersection of AI and the physical world
At N47, we've spent years backing builders who master the full stack. Verkada, Skydio, Tractian. These aren't software-only companies. They build complete systems where hardware, firmware, and software work together to solve real problems in the physical world.
The intersection of hardware and software is where we've always focused, and it's where we believe the most transformative companies emerge. Now we're watching something bigger take shape.
Machine intelligence hit an inflection point.
Vision models can see. Neural networks can predict. LLMs can reason. Our portfolio companies are already at the cutting edge, applying these capabilities to problems that exist outside screens. The results are remarkable.
But AI alone isn't enough.
To truly sense, understand, and act in physical space, you need infrastructure. The new electric stack is that infrastructure, and it's the foundation for what comes next.
When you combine advances in machine intelligence with mastery of batteries, motors, power electronics, and embedded compute, you unlock entirely new categories of possibility.
This is Physical AI.
When I first read "The Datacenter as a Computer" by Luiz Barroso and Urs Hölzle, it reframed my understanding of cloud computing. Recently, Urs highlighted The Electric Slide by Packy McCormick and Sam D'Amico. And their work captures something essential about how the physical world is being rewired. This piece is an extension of these ideas.
Below, I’ll explore what the new electric stack is, why it's foundational to Physical AI, and why mastering both will define the next generation of category-creating companies.
If you're building at this intersection, this is essential groundwork. It's also why we believe N47 is uniquely positioned to partner with you.
What is the new electric stack?
The cloud stack abstracted and scaled intelligence in the digital world. The electric stack now does the same for the physical world.
The electric stack has four layers:
- Lithium-ion batteries now deliver the power density that makes portable flight possible and the energy capacity that makes long-range EVs practical.
- Magnets and motors convert energy into motion with near-perfect efficiency. Neodymium magnets generate magnetic fields strong enough for drones to fight gravity and robots to lift loads, all in palm-sized packages.
- Power electronics control energy flow with precision. IGBTs and SiC MOSFETs switch power thousands of times per second, turning a battery’s DC into precise AC waveforms that spin electric motors.
- Embedded compute coordinates everything. Microcontrollers and DSPs read sensors, run control algorithms, and tell the power electronics what to do.
Each layer—batteries, motors, power electronics, and embedded compute—has seen a 98–99% cost reduction over recent decades, making electricity programmable and action abundant.
Since 1990, the cost of building an electric product has dropped roughly 99%. An average of 12% per year. This is the Electric Slide, i.e., the set of learning curves reshaping what's possible in the physical world.
Why it matters
Every industrial revolution begins with a new way to turn energy into action.
This next wave of innovation comes from connecting intelligence with action. That connection happens through the electric stack. It's what allows AI to reach beyond the screen and fly, drive, grasp, lift, sense, and build. AI controlling electric platforms (e.g., drones, robots, energy grids) gives software the power to shape the physical world directly.
AI models process information. Electric systems move the world. Together, they create intelligent machines that sense, reason, and act. Robots that assemble satellites. Drones that deliver medicine. Equipment that optimizes its own performance in real-time. Systems that balance energy, motion, and decision-making faster than any human could. This isn't theoretical. It's happening now.
The relationship is symbiotic. AI optimizes how energy and motion are used. The electric stack gives AI a physical body. Together, they unlock entirely new categories of intelligent machines—systems that both think and act.
Bits, atoms, and the builders who bridge them
The next wave of progress will come from understanding how each layer of the stack fits together. How scale drives learning. How demand drives innovation.
In early cloud computing, the winners didn't just build infrastructure. They built tools and abstractions that made the cloud accessible to everyone.
In the electric era, the same dynamic is emerging. The winners will build the platforms and systems that make the physical world programmable. We're seeing the step functions already. Each advance in model capability, each improvement in battery density, each breakthrough in power electronics unlocks new applications that weren't economically feasible months earlier.
The pace is accelerating.
The next generation of iconic founders won't choose between hardware and software. They'll master both. They'll be fluent in code and current, in bits and atoms. Understanding both sides is imperative.
This is a call to builders who see the world not as software or hardware, but as both; who design systems that think, sense, and act.
Conclusion
The future belongs to builders fluent in both software and the physical systems it commands.
At N47, we believe the most transformative innovation happens where bits meet atoms. Where digital intelligence connects with the physical world.
Companies across the N47 portfolio are at the cutting edge of this convergence. Verkada applies vision models and neural networks to physical security. Skydio uses AI for autonomous flight in complex environments. Tractian brings machine intelligence to industrial equipment. These teams are proving what's possible when you combine deep AI expertise with mastery of the electric stack. They're creating entirely new categories, not just improving existing ones.
These are the builders we're drawn to. Founders fluent in both code and current, designing systems that connect the digital and the material. Who see the next wave of innovation not just in pixels, but in the physical systems that power our world.
They're proving that the most profound innovation happens at the intersection—where software meets hardware, where intelligence meets motion.
If you're building here, if you're creating products that bridge intelligence and the physical world, we should talk. We recognize these patterns. We understand the tradeoffs. And we're among the few firms fluent in both sides of this equation.
