Welcome back to The Cap Table Newsletter! This bi-weekly newsletter shares key insights on angel investing, start-ups, and investment opportunities.
For the last few years, AI mostly lived on screens. Chatbots. Copilots. Image tools. Workflow software.
But a different category of AI is now pulling in serious capital.
AI that lives in trucks, factories, warehouses, and construction sites.
This week, we’re talking about the rise of Physical AI.
If you want to start angel investing, apply to our syndicate!
Follow our new and improved Instagram to hear about Trending Deals, Founder Stories, & Investor News!
Subscribe Now to get this newsletter delivered straight to your inbox every other Wednesday morning!
The Cap Table is powered by Otter.ai.
Most AI tools fail because they rely on habits users don’t maintain. You have to remember to open the app, invite the bot, and hope everything works. Otter took the opposite approach. With its desktop app, meetings are captured directly from your laptop. No bot to invite. No permissions to manage. No extra step to remember. You open your computer, take the call, and the knowledge is recorded automatically. That’s what makes Otter feel less like a tool and more like infrastructure. Usage compounds because it’s always on, not because you remembered to turn it on.
Try Otter’s new desktop app and see how default on changes everything.
AI money is still massive, but the focus is shifting
AI continues to absorb an outsized share of global venture funding. Roughly half of all venture dollars now flow into AI-related companies.
What’s changing is where that capital is going.
Pure software categories are saturating fast. Competition is brutal. Features get copied. Models improve overnight.
As a result, capital is moving toward areas where AI is harder to replicate and harder to displace.
Physical AI sits squarely in that bucket.
What makes Physical AI different
Physical AI is not software plus hardware.
It is AI operating in environments where failure has real consequences.
Safety matters. Reliability matters. Downtime costs money. Data is expensive to collect. Deployment is slow and unforgiving.
Those constraints change everything.
Iteration slows down, but defensibility increases. Moats are built through deployment, not just code. Once these systems work, they are very hard to rip out.
Three companies defining the Physical AI category
Bedrock Robotics
Founded by Boris Sofman, former autonomy lead at Waymo and former CEO of Anki
Focused on retrofitting autonomy onto existing heavy construction equipment already operating on job sites
Raised $270M in a 2026 Series B, bringing total funding to ~$350M
Valued at $1.75B less than two years after founding
Targeting real production deployment in construction, not pilots or demos
Skild AI
Founded by Deepak Pathak and Abhinav Gupta, leading researchers in robotics and reinforcement learning
Building a general-purpose foundation model for robotics that transfers across machines and tasks Raised $1.4B in a 2026 Series C, following a $300M Series A and $500M Series B
Valued at over $14B in 2026
One of the largest financings ever in embodied intelligence and robotics
Waabi
Founded by Raquel Urtasun, former head of Uber ATG and longtime autonomy researcher
Building autonomous trucking using a simulation-first approach
Raised $1B+ in its latest 2026 round, bringing total capital raised to over $1.2B
Focused on freight autonomy, where small efficiency gains compound across massive supply chains
Why this is happening now
Physical AI has always been the bigger prize, but it needed several things to converge at once.
Models that generalize better. Simulation that is good enough to replace years of field testing. Cheaper sensors and edge compute. Real economic pressure from labor shortages and productivity ceilings.
Those pieces are finally aligning.
What used to be a research project is becoming an operating system for real industries.
Why investors are leaning in
Physical AI changes the risk profile.
Software AI is fast, but fragile. Distribution is cheap. Capabilities diffuse quickly.
Physical AI has friction, and that friction becomes defensibility.
Proprietary real-world data. Long deployment cycles that lock in customers. Safety and reliability credibility that takes years to earn. Contracts tied to operations, not experimentation.
These companies are selling into the physical economy: transportation, infrastructure, manufacturing, logistics.
Those markets are measured in trillions, not features.
The world we are walking into
If Physical AI works, the impact won’t feel flashy. It will feel structural.
Freight becomes more predictable. Construction timelines compress. Factories and warehouses become software-defined. Robots shift from single-purpose machines to flexible systems.
The work itself changes. Fewer repetitive tasks. More oversight, maintenance, and exception handling.
This transition won’t be even. Industries under the most margin and labor pressure will adopt first. Others will follow when they have no choice.
My Take
The first wave of AI was about digital productivity.
The next wave is about physical leverage.
Physical AI will produce fewer breakout companies than software AI, but the winners will be harder to replace and more deeply embedded into the economy.
It is slower. It is harder. It is messier.
That’s exactly why it’s becoming its own sector inside AI, and why capital is showing up in size.
Until next week, Elana ✌️
Resources
If you enjoyed this week’s newsletter - feel free to check out some of our past articles:
👋 That’s all for now friends! See you next week.
In the meantime, Follow our instagram to see the latest founder and VC updates.
💌 Join our subscribers and sign up for this weekly Cap Table Newsletter if you haven’t already!
Also if you are interested in starting to Angel Invest you can apply to our syndicate to see our weekly deal flow!
Disclaimer: The Cap Table DOES NOT provide financial advice. All content is for informational purposes only. The Cap Table is not a registered investment, legal, or tax advisor or a broker/dealer.