RobustAI Named Physical AI Startup to Watch
Being on AIM Research’s list of noteworthy Physical AI Startup to watch both reinforces growing interest in AI’s more tangible applications and the importance of our mission to make robots work for humans. But with this momentum brings up some questions around physical AI that could use some attention and explanation.
How much do people actually know about physical AI?
How do the gen AI tools of the moment - the Geminis, ChatGPTs, Copilots, and Einsteins of the world - influence perception, understanding and curiosity around other types of AI?
What can physical AI companies do to help promote understanding, communicate value and put the right context around their solutions?
Physical is just different.
Earlier this year at the Robotics Summit in Boston, Teradyne Group President, Ujjwal Kumar gave a keynote on ‘Transformation of the Robotics Industry.’ When he discussed Digital Al vs. Physical AI, in terms of impact, he said,
“There are different stakes in physical space…this is why physical AI - and what Amazon and NVIDIA have done - is so important. Physical AI is what will drive real transformation.”
Without a doubt, these two industry giants have certainly driven the market and shown the world the potential and depth of applications for physical AI. But the empirical statement in this quote is actually the first part. The different stakes of physical space are the challenge and the opportunity for AI.
The basics: mass, volume, area, perimeter, height, width, and radius never lie and can’t be understated. These spatial points of reference are what people use to sense, experience, and adapt to the world around them. The ways in which AI activates our tangible world is what changes everything.
A robot is “an AI,” right? Mmm, not quite.
Syntax is an enemy of the state when it comes to helping people distinguish how technologies work together vs. being synonymous. A Forbes article suggests,
“...even industry insiders find themselves in a perpetual game of catch-up. From autonomous vehicles to robotics, each new application seems to spawn its own terminology, leaving many confused and pondering a deceptively simple question: Just how many flavors of AI are there?"
Whether they admit it or not, a lot of people - technically-minded or not - walk up to the AI painting and are left thinking “It’s cool, but what am I looking at?” Software, hardware, software fueling hardware - AI is its own ecosystem full of coexisting, independent, and homeostatic technologies. The important exercise is not to overwhelm yourself trying to sort it all out - it’s to be motivated by and hopeful about what it can do to make life better for people.
So, to answer the question, no. A robot is not an AI. It’s a form physical AI can take and enable to solve real world problems. Not too shabby.
But gen AI is all everyone is talking about…for now.
It’s been a couple of years now, but there continues to be a LOT of discussion and excitement around gen AI - sort of like the current humanoid movement in the robotics space (except a lot more accessible). It wouldn’t be fair or accurate to say these things are just having their 15 minutes, but it would be reasonable to suggest that there’s a lot more going on behind the scenes in AI and robotics that have substantial implications.
In some ways, what’s being talked about right now is scratching the surface of what is to come. That’s the “waiting in the wings” feeling around physical AI. Similar principles, bigger punch. As Kumar suggested in his keynote, outcomes with physical AI will be more transformative - and nothing drives the market more than proof. Outputs of words, visuals, concepts, and information are fascinating, but still have digital limitations.
Physical AI companies need to lead with proof and purpose.
Simply put, physical AI, when done right, is meant to work with and for people - but it’s not simple. In the case of physical AI and robotics, our Director of Engineering, Benjie Holson, aptly summarizes, “robots are hard because the world is complicated.” (Check out his article on that topic). So how do leading physical AI companies communicate the value and vision without adding to the noise or confusion? With physical proof and accessible applications.
If a solution exists in an accessible, relatable space, it’s immediately more compelling. There’s a reason why robotics and automation goes absolutely bonkers for industry tradeshows. Everyone wants to see, touch, hear, and be delighted by something new and impressive to behold. But when it comes to actually investing in solutions, there’s a difference between putting on a show and proving success. You have to check both of those boxes these days.
For our team at RobustAI, we see the opportunity for collaborative productivity in physical spaces, using AI to bring the best of robots and humans together in common physical spaces.The more physical AI companies can show business outcomes and communicate the purpose behind them in ways their audiences - and the general public - can understand and appreciate, the more AI, as a whole, becomes mainstay over magic.
