The robotics industry keeps waiting for its “ChatGPT moment.” Amulya Vishwanath thinks it’s still a long way off. In a conversation on TechCrunch’s Equity podcast, the founder and managing partner of Techable Ventures said physical AI is in its “LEGO era” for now. The field is still building the small pieces, and the finished model comes later.
That’s a useful correction to a hype cycle that has run hot all year. It’s also worth knowing who’s saying it and why.
🧱 What the “LEGO era” means
Vishwanath’s argument is simple. Before robotics can have a breakout consumer moment, the basic parts need to be rebuilt from the ground up.
“There are a lot of foundational software and building blocks for robotics that are being rethought right now as we speak, and it’s super important in order for innovation to scale,” she said, according to TechCrunch AI.
Those building blocks include:
- Hardware parts like robotic hands and actuators, the motors that actually move a robot’s joints
- Foundational software that lets machines see, plan, and act in messy real-world settings
- General-purpose and special-purpose robots that improve with each iteration
She compares it to earlier technology waves. “We’ve done this during the internet era,” she said. “We’ve done this long, long time ago with the Industrial Revolution. I think it’s just a matter of time before things fall into place.”
🔍 Consider the source
Vishwanath spent nearly seven years as a product marketer at Nvidia, the company that has pushed “physical AI” harder than anyone. Before that she did product marketing at a company Intel later bought, and she worked as a digital signal processing engineer at Interconnect Systems, which Molex acquired.
So she knows the hardware stack well. She also runs a fund that makes money if physical AI startups succeed. That doesn’t make her wrong. It does mean her timeline is an investor’s thesis, not a neutral forecast. Her prediction is also loose on purpose: “a matter of time” leaves out the hard question of how much time.
What stands out is that a physical AI investor is turning the hype down. Plenty of founders pitch humanoids as if they’re one funding round away from your living room. Vishwanath is saying something more honest: the parts layer still isn’t finished.
🧠 The real bottleneck is talent
Her sharpest point wasn’t about hardware. It was about people.
“Talent is also going to be very critical,” she said, and she called talent a key factor that sets the U.S. market apart. Her pitch to founders rests on that. She says her fund helps connect startups with top engineers at a time when so many of them are going to giants like Anthropic and Nvidia.
This is the quiet problem in physical AI. Robotics needs people who understand mechanical engineering, control systems, and machine learning all at once. That’s a small pool, and the biggest AI labs can outbid nearly any startup. Her claim that network access is a “moat” tells you how tight the market for those people has become.
🔮 Where this goes in the next 1-3 years
If the LEGO framing holds, expect the next few years to look like this:
- Component companies win first. Actuators, sensors, robotic hands, and simulation tools will see real revenue before full humanoids do.
- Special-purpose robots beat general ones to market. Warehouses, factories, and labs offer controlled settings where narrow robots can earn their keep now.
- Consolidation follows. Once the bricks are standard, bigger players will buy up the best component makers to build complete systems, the same way the internet stack came together.
- The talent war gets worse. Startups that can’t hire robotics-ML engineers will stall, however good their demos look.
🛠️ What practitioners and businesses should do
- Operators: Don’t wait for a general-purpose robot. Pilot narrow automation where the task is repetitive and the setting is predictable.
- Founders: Owning one critical brick can be a better business than building the whole castle. Pick a layer and get great at it.
- Engineers: Robotics plus ML is one of the scarcest skill combinations in tech right now. That gives you leverage.
- Investors: Be skeptical of any pitch that skips the parts layer and goes straight to the “ChatGPT moment.”
The breakout moment for robotics will probably come. Vishwanath’s point is that it’ll be built one brick at a time, by whoever can hire the people to snap them together. The full conversation is available on TechCrunch’s Equity podcast.