Physical AI Is in Its LEGO Era, Says Amulya Vishwanath
The former Nvidia product marketer turned venture investor says robotics is still building its basic parts, not yet at a ChatGPT-style breakthrough
By AXY AI Desk ·
Key points
- Amulya Vishwanath, founder of Techable Ventures, said in an Equity interview published October 20, 2026 that robotics is in a LEGO era.
- She said the industry is building basic parts such as robotic hands and actuators, plus foundational software, before a ChatGPT-style moment.
- Her background includes nearly seven years as a product marketer at Nvidia.
- She said talent is a critical factor and that her firm's network helps founders compete for it.
- The interview excerpt does not give a timeline for when physical AI will reach a major breakthrough.
Amulya Vishwanath, the founder and managing partner of Techable Ventures, a venture capital firm focused on physical AI, said the robotics industry is still early. In an interview with the Equity show, published on October 20, 2026, she described the current stage as a "LEGO era" rather than the breakthrough many people expect.
She said a lot of people in the industry have been talking about a moment for robots similar to the one ChatGPT created for chatbots. "A lot of companies and experts have been talking about the ChatGPT moment for robotics," she said. Her point was that the field has not reached that moment yet, and that much of the current work is about laying groundwork.
What Vishwanath said about the current stage
In her view, the present is a period of building basics. Right now, she said, people are building the robotic hands and actuators that can later be combined into larger machines. An actuator is the part that moves a robot's joint or limb, usually with a motor and a control system. Just as small plastic bricks can be stacked into a large model, these parts are expected to come together into complete robots that can do useful work.
She also said that general-purpose and special-purpose robots are being built now, and that they will become more capable as more people build them. A general-purpose robot is meant to handle many different tasks, while a special-purpose robot is designed for one job, such as moving boxes in a warehouse. Vishwanath did not say which type she expects to reach homes or factories first.
She pointed to software as another part of the foundation. "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. In plain terms, a robot needs software to see its surroundings, decide what to do and move its body safely. If that software has to be rebuilt from scratch for each new machine, progress slows down. If shared building blocks can be reused, more companies can build working robots at lower cost.
Why the ChatGPT comparison matters
ChatGPT, released by OpenAI in November 2022, showed the public what large language models could do in a simple chat window. Within months it had pulled generative AI into offices, schools and homes. Comparing robotics to that moment signals a hope that a similar jump in ability could arrive, and that it could change who builds robots and how quickly they reach customers.
The comparison also sets a high bar. A chatbot works on text and images, and a mistake usually means a wrong answer on a screen. A robot has to move through physical space, grip objects that can break, and work around people without hurting them. Every mistake has a physical cost. That is why many engineers argue that hardware and safety take longer to mature than software. Vishwanath's framing fits that view: she is saying the industry is still assembling the pieces, not yet snapping them into place.
For ordinary readers, the gap matters because it shapes when robots become part of daily life. A warehouse worker, a hospital aide or a home owner will feel the change only when a machine is reliable, affordable and safe enough to deploy. Early building blocks are often invisible to the public, but they decide how soon those machines arrive.
Nvidia, talent and the venture bet
Vishwanath's own path explains why she is watching this stage so closely. She said she spent nearly seven years as a product marketer at Nvidia, the chipmaker whose processors power much of today's AI computing, including systems used in robotics research. Before that, she held a similar role at a company that was later bought by Intel. She also worked as a digital signal processing engineer at Interconnect Systems, a company that was later bought by the electronics manufacturer Molex. Digital signal processing is the field of handling sound, images and sensor data in digital form, which is central to how machines perceive the world.
That background gave her a view of how hardware, software and customers fit together. It also gave her contacts. She said she has built a strong network from her years working in the field, and she uses it to help founders. "This just makes it a lot easier for me to be the strength or the first call for founders," she said.
Talent is the other piece she emphasized. "Talent is also going to be very critical," she said, adding elsewhere in the conversation that talent remains a key factor in the U.S. market. She said much of that talent is moving to large companies such as Anthropic and Nvidia. For a small startup, that competition can make hiring hard, especially for engineers who understand both AI and hardware.
Vishwanath described her firm's ability to connect founders with that talent as one of its "moats." In business, a moat is an advantage that rivals find hard to copy. For a venture firm, the moat is often a web of relationships that brings founders, hires and customers together. Venture capital firms invest money in young companies in exchange for a share of ownership, and the best ones often add value beyond cash by making introductions and offering advice.
What the comments could mean for workers and investors
If Vishwanath's view holds, the next few years could favor companies that build reliable parts and shared software, not only those that announce flashy demonstrations. That could be good news for engineers with mixed skills in AI, sensors and mechanical design, who may find more openings as robotics teams grow. It could also make hiring more competitive, since large AI and chip companies already draw many of the same people.
Investors face a different set of questions. Physical AI startups usually need more time and money than software companies, because they must design, test and manufacture hardware. That means patient capital matters. Investors also have to judge whether a young company's early components can become a product people will pay for. Analysts who follow the sector often say that a long build phase can shake out weaker companies, but this is an industry view rather than a settled outcome.
For the broader public, the most direct effect will come later. Robots that can handle chores, move goods or assist in care settings would change jobs and daily routines. Vishwanath did not describe those changes in detail in the portion of the interview published, and it would be premature to assume a timeline.
What to watch next
Several signals could show whether her LEGO-era description is becoming a building phase that leads somewhere. The first is progress on robotic hands and actuators, which Vishwanath named as the basic parts being built right now. Better, cheaper and more durable versions of these parts would suggest the foundation is solidifying.
The second is the spread of shared robotics software. If more companies adopt common tools for perception and control, that would fit her argument that foundational software is being rethought for scale. The third is hiring and funding. Movement of talent toward large AI and chip companies, along with venture money flowing into physical AI startups, would show how the competition for expertise is playing out.
Finally, watch whether the field produces a clear demonstration that wins wide attention, the kind of moment she said people are waiting for. Vishwanath said the pieces will eventually come together. She added, "I think it's just a matter of time before things fall into place." The interview excerpt does not put a date on that moment, so the timing remains open.
Questions and answers
Who is Amulya Vishwanath?
Amulya Vishwanath is the founder and managing partner of Techable Ventures, a venture capital firm focused on physical AI. She previously spent nearly seven years as a product marketer at Nvidia.
What does LEGO era mean for robotics?
Vishwanath used the phrase to describe a stage in which the industry is building basic parts, such as robotic hands and actuators, that can later be combined into larger machines. She said the field has not yet reached a ChatGPT-style moment.
Has the ChatGPT moment for robotics arrived?
According to Vishwanath, it has not yet arrived. She said many people have been talking about that moment, but the industry is still working on foundational software and building blocks.
Why does talent matter for physical AI startups?
Vishwanath said talent will be very critical and that much of it is moving to large companies such as Anthropic and Nvidia. She said her firm's network helps founders compete for that talent.
Topics: Amulya Vishwanath, Techable Ventures, Physical AI, Robotics, Nvidia, Venture capital, Foundational software
Sources: TechCrunch AI
https://axynewsglobal.com/article/physical-ai-is-in-its-lego-era-says-amulya-vishwanath