4 Wheels Good, 2 Legs Best

In the past year, there's been an absolute explosion in commercial, industrial, and even hobbyist robots being released. These 'bots have been coming from companies ranging from well-established titans of the industry, like Milestone or Silver Swan, to brand new independent startups, creating robots for specific niches.

Despite this Cambrian explosion of new robots, one thing is startlingly common among them: They all, for the most part, are humanoid.

In the past, robots traveled fairly strictly by wheel, tread, sometimes flight - but rarely legs. Sure, you had quadropeds like Boston Dynamic's Spot 3.0, and their line of bipedal robots, but these were incredibly expensive machines largely meant for research and not commercialization. Now, robots with movement as fluid as Atlas are becoming more and more common.

The reason for this, says roboticist Sophie McClelland, is due to a fundemantal shift in how these machines are programmed.

"In the past, software engineers would need to spend years developing advanced kinematic models to tell these machines how to walk, which was an incredibly difficult task. Nowadays, every robot just uses some variant of the openAware framework. Because the developers built openAware using brain mappings from adult humans, some knowledge of 'walking' is inherently baked into the software, and most of the hard work of telling a robot how to walk can be skipped entierly. It's truly astounding."

McCelland also pointed out that not only are bipedal/humanoid robots easier to make, but openAware might only be compatible with them to begin with.

"We've run experiments on using openAware in wheeled robots since - despite being better suited to human environments, bipeds are ultimately more expensive to manufacture. However, the robots we developed with these systems required significantly more work to make them cooperate with their means of locomotion. The problem with relying on openAware is, as incredible as it is, we still don't fully understand it - it's as though we put a brain in a jar and plugged it into a machine, but we've done away with all the dirty anatomy and reduced it to code anyone can download"

We have yet to see the long-term effects of this new innovation. Many have complained of current machines suffering from the 'uncanny valley' effect, while others are concerned about the ethics of open-sourcing the ability to create consciousness. Skeptics argue this technology will be another short lived fad, joining the likes of NFTs and Large Language Models of yesteryear. Time can only tell, but the sudden prevalence of these machines in our daily lives seems to suggest they're not going away any time soon.