AI-powered humanoid robots are closer than we think, says Nvidia exec

The key to AI deployment in the modern world is robotics, claims one Nvidia exec who envisions humanoid machines walking alongside us in the not-too-distant future.

Speaking at the recent 2024 World Congress on Innovation and Technology event in Armenia, Rev Lebaredian, VP of Omniverse and Simulation Technology at Nvidia, predicted that, like electric cars, AI-powered humanoid robots will soon become affordable enough to reside in everyday homes and workplaces.

“They have to be humanoid, or humanoid-like,” he added.

Robots at work

In his talk, Lebaredian explained the humanoid form factor is crucial to the successful deployment of robotics in order for them to coexist in a world built exclusively for humans.

The rationale behind this concept is rooted in the physical design of the world around us – factories, warehouses, offices, hospitals and other public spaces are all structured for human movement and activity. Rather than rebuilding a world suitable for computers, Lebaredian says that processor-brained beings should instead be tailored to existing infrastructure.

While other types of specialized robots will still have their place, humanoids will offer the versatility to perform a wide range of tasks already optimized for human use, he added, noting some human-like features may not be necessary, for example a robot operating in a single-storey factory may be more cost-effective when built with wheels rather than legs, but human-like movement still plays an important role.

For example, humanoids could stock shelves in a retail store or handle logistics in a warehouse.

Combining robotics with artificial intelligence

Technological advancements have already made humanoids feasible today, however the additional layer that artificial intelligence provides is set to accelerate their usefulness.

Speaking about challenges like a robot’s ability to balance, walk and navigate, Lebaredian added: “It’s only recently, because of these advances in AI and computing, we actually see a line of sight to solving all of them.”

An AI-powered ‘brain’, which Lebaredian described as the, “missing ingredient for creating a useful general purpose robot,” together with the ability to learn from simulations before being deployed and the ability to interact with their surroundings are among the key factors characterizing upcoming robots.

Through Nvidia’s Omniverse platform, such robots are already being trained in highly accurate simulations before being deployed to the real world, where they continue to learn and adapt.

Moreover, the cost associated with developing these humanoids is expected to drop rapidly. Just as smartphones became more affordable, humanoid robots are likely to follow a similar trajectory, which will enable them to be deployed across more industries.

“I believe that they’re going to become so common, they’ll be like our cell phones. They’re going to get cheaper and cheaper very, very quickly,” Lebaredian added, noting this dystopian-like image could become a reality in a matter of years, not decades.

Humanoid implications for society

The deployment of such digital beings could have far-reaching implications, for example in healthcare, humanoids could assist in surgeries where they are more adept at fine motor movements.

In logistics and manufacturing, they could fill critical gaps left by labor shortages, adding to the automation that has already helped companies keep up with demand and competition.

In the two years that have followed the public preview launch of ChatGPT, and the subsequent AI boom, we’ve already seen businesses keenly adopt AI in the hope of realizing productivity and efficiency gains. Lebaredian hopes this eagerness will continue as we shift towards robotics and AI in physical environments.

Summary

If for nothing else at all, generative AI has proven that we are capable of creating solutions to tackle societal struggles, like dipped productivity and talent shortages.

The ability for humanoid robots to integrate seamlessly into human-built environments and perform a wide variety of tasks represents the next wave of innovation in an increasingly automated world.

Furthermore, as the cost of developing these robots declines, Lebaredian predicts we will soon be paying little more than the actual manufacturing and material costs, fueling a surge in their adoption across various industries and, before long, in our homes.