OpenAI and Figure join the race to humanoid robot workers

The jarring emergence of ChatGPT has made it clear: AIs are advancing at a wild and accelerating pace, and they're beginning to transform industries based around desk jobs that typically marshall human intelligence. They'll begin taking over portions of many white-collar jobs in the coming years, leading initially to huge increases in productivity, and eventually, many believe, to huge increases in unemployment.

If you're coming out of school right now and looking to be useful, blue collar work involving actual physical labor might be a better bet than anything that'd put you behind a desk.

But on the other hand, it's starting to look like a general-purpose humanoid robot worker might be closer than anyone thinks, imbued with light-speed, swarm-based learning capabilities to go along with GPT-version-X communication abilities, a whole internet's worth of knowledge, and whatever physical attributes you need for a given job.

Such humanoids will begin as dumbass job-site apprentices with zero common sense, but they'll learn – at a frightening pace, if the last few months in AI has been any kind of indication. They'll be available 24/7, power sources permitting, gradually expanding their capabilities and overcoming their limitations until they begin displacing humans. They could potentially crash the cost of labor, leading to enormous gains in productivity – and a fundamental upheaval of the blue-collar labor market at a size and scale limited mainly by manufacturing, materials, and what kinds of jobs they're capable of taking over.

The best-known humanoid to date has been Atlas, by Boston Dynamics. Atlas has shown impressive progress over the last 10 years. But Atlas is a research platform, and Boston Dynamics founder Marc Raibert has been clear that the company is in no rush to take humanoids to the mass market.

"We have to remove the pressure to make things more reliable, more manufacturable, and cheaper in the short term," he told IEE Spectrum last year. "Those are things that are important, but they’re in the way of trying new things. The pitch I made to [parent company] Hyundai explicitly says that, and proposes funding that extends long enough that we’re not distracted in the short term."

Hanson Robotics and Engineered Arts, among others, have focused on human/robot interaction, concentrating on making humanoids like Sophia and Ameca, with lifelike and expressive faces. Ameca in particular communicates using the GPT language model.

Others are absolutely focused on getting these things into mass usage as quickly as possible. Elon Musk announced Tesla's entry into humanoid robotics in 2021, and the company appears to have made pretty decent progress on the Optimus prototypes for its Tesla Bot in the intervening months.

"It's probably the least understood or appreciated part of what we're doing at Tesla," he told investors last month. "But it will probably be worth significantly more than the car side of things long-term."

Just this year, entrepreneur Brett Adcock, founder of the Vettery "online talent marketplace," and more recently Archer Aviation, one of the leading contenders in the emerging electric VTOL aircraft movement, announced his latest venture is focused on humanoid robots.

The first phase of a Musk-like "master plan" for the new company, Figure, aims first to "build a feature-complete electromechanical humanoid," then to "perform human-like maipulation," then to "integrate humanoids into the labor force."

Figure aims to create "the world’s first commercially viable general purpose humanoid robot" – at this stage called the Figure 01. This fella will stand 5 foot 6 inches (168 cm) tall and weigh 132 lb (60 kg) according to the company. It'll be able to lift a targeted 44 lb (20 kg) of "payload," walk at speeds up to 1.2 m/s (2.7 mph/4.3 km/h), and operate for up to five hours on a charge.

As a two-legged humanoid, it'll be able to climb stairs, walk around, and generally operate in most of the environments humans can – at least, eventually. With humanlike hands and body mechanics, it should be able to use the tools we use, and do a variety of our jobs.

Adcock is approaching this project with trademark speed and aggression. When Archer launched, it hired a bunch of talent away from other leading eVTOL companies and made very quick progress, both technically and in terms of fundraising. Indeed, according to the AAM Reality Index, Archer is now the third-best-funded eVTOL company in the world, and looks likely to have aircraft in service in 2025, around the same time as Joby Aviation. Joby was founded in 2009, and Archer came out of stealth in 2020.

Figure appears to be taking the same approach; according to an interview with IEEE Spectrum, the company had more than 40 engineers on board at the beginning of March, with key talent from the Florida Institute for Human and Machine Cognition, Boston Dynamics, Tesla, Waymo and Google X, among others. So while the company doesn't seem to have anything actually built yet, it's a serious venture proceeding at top speed.

"We face high risk and extremely low chances of success," reads the Figure website. "However, if we are successful, we have the potential to positively impact humanity and to build the largest company on the planet."

Meanwhile, OpenAI has signaled it's getting interested in humanoids again as well. OpenAI wants to create an "artificial general intelligence" or AGI – a machine that can do nearly all tasks better than a human. Through the GPT language model, the company expects to knock a lot off that list in the short term future, but it'll need to put its intelligence in a body to handle the rest.

OpenAI had its own robotics division for many years, and indeed built a humanoid hand capable of fine manipulation and sensing that used neural networks and reinforcement learning to figure out how to solve a Rubik's cube, one-handed.

But the company shut down its robotics team in 2021, telling VentureBeat that it was focusing its efforts in the area where it was progressing fastest – generative AI and large language models like GPT. The key difference, according to co-founder Wojciech Zaremba, is data and computing power. The AI-powered robots proved capable of learning physical tasks "extremely well" – there's just far more text training data than robot-relevant video data, and text is much faster for a neural model to process.

So text is where they focused their efforts, and the results are hard to argue with; OpenAI now has the fastest-growing platform in history, a game-changing breakthrough of the highest order in the GPT language model.

But through an investment in Norwegian company 1X, formerly known as Halodi Robotics, it seems OpenAI is ready to get back into embodied intelligence. Details are scant at this point; 1X announced the close of a US$23.5 million Series A2 funding round on March 23rd, led by the OpenAI Startup Fund.

Halodi made a certain amount of progress with its Eve robot, a humanoid on a wheeled base, a version of which can be seen in the video above. But a wheeled robot is a limited one, in terms of the human tasks it can take over, and 1X is now focused on a bipedal platform called Neo, to "explore how artificial intelligence can take form in a human-like body."

It looks somewhat similar to Figure's design, in that it appears to use electric actuators rather than hydraulics, it runs a screen for a face, and it uses humanoid hands.

The company has little more to say at this point, other than the words "Summer 2023," which appears to indicate we'll be meeting Neo-on-legs very soon. It'll be fascinating to see what OpenAI's contributions will bring to a project like this.

This space could well become the next high-tech gold rush, since as both Tesla and Figure point out, whoever gets a general-purpose humanoid out there first, that's capable of doing real work across a decently broad set of tasks, is in an amazing position to generate near-unlimited money and value.

Workers tend to be much cheaper to buy than to hire. Some of the most shameful chapters in human history demonstrate how much wealth can be generated if you own a subservient, non-unionized workforce that operates under zero OHS requirements, requires no pay or sick leave, and can be told to handle a range of different tasks.

A humanoid robotics revolution could deliver a lot of the same benefits in a much more morally defensible way – although if it's successful enough (and it seems it will be, it's just a matter of when) – it'll open its own Pandora's box of societal consequences.

Of course, the human form certainly isn't optimized to get today's work done, so at some stage it'll start making sense to create superhuman robots that are faster and stronger than the best of us, with greater reach, extra legs and arms, and completely different form factors. And that's when things will start getting really weird and scary.

Sources: 1X, Figure