The Secret to Manifested AGI

Something is up at OpenAI…

And it’s not a small development.

The company has been actively recruiting top researchers to develop the artificial intelligence (AI) needed for humanoid robots – an application that I refer to as manifested AI.

It’s not just the high-profile hires that signal something is different. OpenAI is clearly building out a large team to tackle the problem.

Just have a look at the key positions that OpenAI is looking to fill in its robotics division. These aren’t just four individual positions. These are entire job categories where they need a large number of engineers.

Robotics Job Postings at OpenAI | Source: OpenAI

Each job description leads with a description of the purpose of OpenAI’s robotics team:

Our Robotics team is focused on unlocking general-purpose robotics and pushing towards AGI-level intelligence in dynamic, real-world settings. Working across the entire model stack, we integrate cutting-edge hardware and software to explore a broad range of robotic form factors. We strive to seamlessly blend high-level AI capabilities with the constraints of physical systems to improve people’s lives.

It goes without saying that these aren’t low-level jobs. The mechanical engineering positions have a salary range of $380K–460K a year, plus equity in OpenAI.

Naturally, they are looking for new members of the robotics team with specific skill sets and experience if they are offering those kinds of compensation packages.

Perhaps the most interesting signal I found was in the job descriptions for the mechanical engineering positions:

Experience designing mechanical systems intended for high volume (1M+).

The hint is in the “high volume (1M+).” It’s a reference to manufacturing more than 1 million mechanical systems… a year.

This clearly isn’t a pet project. This isn’t some kind of research lab to support OpenAI’s pursuit of artificial general intelligence (AGI).

OpenAI has entered the race for general-purpose intelligent humanoid robots.

And it is waaaaay behind.

OpenAI Is Playing Catch-Up

As we explored in yesterday’s Bleeding Edge – Musk’s Billion-Dollar Signal, there are a number of companies like Unitree, Tesla, Apptronik, Agility Robotics, Figure AI, 1X Technologies, and others that already have functional commercial products in the early stages of deployment.

These companies have been at it for years. So why now? Why is OpenAI jumping into the race?

After all, OpenAI is a software company. Its frontier AI models are primarily used for things like teaching, medical advice, “how to” questions, information research, writing software code, summary generation, data analysis, and even creative writing.

Those in the media that are aware of these latest developments at OpenAI believe this is just about the money. OpenAI is simply in pursuit of a multi-trillion-dollar market opportunity.

Or maybe this is about the long-running and contentious feud between Elon Musk and Sam Altman, which began after Altman radically shifted the focus of OpenAI away from its original non-profit principles.

But these opinions completely miss the big picture.

For artificial intelligence to achieve general intelligence (AGI) and to be manifested into forms that will be invaluably useful to the human race, these AI models need a whole lot more real-world data, and they need to gather it quickly.

Poptimus? | Source: Tesla

And what better way to collect that data than through the “eyes” and sensors of mobile, autonomous robots? The more of them out there in the field, the more real-world data that can be collected.

x

Tesla’s Edge

This is a key reason why Tesla is making such rapid progress with Optimus. Tesla was able to “seed” Optimus with its autonomous driving software, which was built on billions of miles of real-world driving data, specifically video with a 360-degree perspective around the car.

I’ve been a vocal proponent of vision-based artificial intelligence models since early 2018, when I pounded the table to buy Tesla (TSLA). The stock is up 2,000% since then, up about 21 times since that recommendation.

Back then, Tesla was largely hated by the financial media, proclaimed to be on the verge of bankruptcy, and the vision-based approach to self-driving software was widely believed to be “impossible” by the industry experts who insisted that using LiDAR – stands for Light Detection And Ranging, it’s light-based, remote sensing laser technology – was the only way autonomous driving could ever be achieved.

I had never received so much criticism about my research ever, but I was steadfast in my research and conclusion that Tesla’s approach would be successful and that it is the only way to achieve self-driving at scale.

And now we know, without a doubt, that Tesla’s strategy was not only technically sound but also the most cost-effective and highest-performing model for autonomous robotics. After all, an electric vehicle with self-driving software is nothing more than a robot on four wheels…

Which is why Tesla’s “seed” for Optimus was valuable. It gave Optimus the “intelligence” to navigate the real world, only it wasn’t streets and highways, it was offices and factory floors.

Vision is the key – vision through high-resolution cameras as sensory inputs for learning about and interacting with the real world.

That’s why OpenAI needs to move quickly with robotics.

It needs to collect more real-world data to improve its own AI. It needs this data to achieve AGI. After all, we know that Musk won’t share all of his vision-based data with Altman, and neither will any of OpenAI’s robotics competitors.

Vision is the secret to manifested AGI – our walking, talking, soon-to-be friends and collaborators in this world full of autonomous general intelligence robots that is right around the corner…

Jeff

P.S. Manifested AI is a multi-trillion-dollar market trend in the making, with Musk and his team at Tesla leading the charge. Optimus has been ahead of the competition at every step.

One opportunity for those keen to get ahead of this megatrend lies in the “brain” behind the vision-based artificial intelligence network that powers both the Tesla full self-driving technology as well as empowers Optimus to interact with the real world.

You can go here to learn more about this Manifested AI hardware… and the company supplying it.