"How about this… I can guarantee that by the time we’re done with Amazon, subscribers will have, at a minimum, doubled their money… in one of the largest companies on earth. Today, I’m going to do the same with Tesla. I’m going to show you a company that is grossly misunderstood by the market, and I’m going to show you where the real value is."
— Jeff Brown, The Near Future Report, November 2018
I wrote the above five years ago.
Not only was my predication about Amazon’s share price spot on, but my research on Tesla was incredibly accurate as well.
And ironically, it was by far my most controversial recommendation.
Which makes this week’s developments at Tesla that much more interesting.
These are the three laws of robotics developed by science fiction giant Isaac Asimov.
What’s incredible is that they were written 81 years ago back in 1942. Asimov first presented these three “laws” in a short story called Runaround, and they became further memorialized in Asimov’s Robot series of short stories and novels.
It was prescient for Asimov to think so far into the future about what kind of framework would be needed for employing robots in our daily lives.
For decades, these laws haven’t been needed though. The reality is that the technology simply didn’t exist.
But we need them now.
This week, Tesla announced its “Gen 2” version of its artificial intelligence (AI) powered robot — Optimus.
For those that might want to watch the whole video clip, you can do so right here.
It’s less than two minutes. I suggest watching all the way to the end — it’s worth it.
This topic is a perfect continuation of yesterday’s Outer Limits, when we explored the significance of DeepMind’s latest Gemini AI.
You’ll recall that most popular generative AIs right now — like Open AI’s ChatGPT — are “just” software trained on internet information. They have no ability to “see” or receive data past a specified date… or use data inputs from the present.
DeepMind’s latest Gemini AI, however — released last week — is multimodal. It has the ability to intake sensory information — from different inputs such as images, audio, and video from the real world — and use that information to infer and respond.
Now, let’s put that into context with Tesla’s "Gen 2" Optimus announcement…
It is huge. And it is absolutely real.
There is no computer graphics in the video above — nothing is staged, everything is real-time, and none of the video has been speed up.
The advancements in Optimus are impressive. Some of the highlights include:
This is remarkable. Tesla has built a full-stack humanoid robot.
And it isn’t designed just to perform a small number of pre-programmed tasks. Elon Musk and his team are building it to do pretty much any tasks that humans don’t want to do.
In other words, it is a general purpose robot. I’m sure we can all think of at least a dozen tasks we’d love to hand over to something like Optimus.
Well, our wish is about to come true…
Even more exciting is that Musk is quoted as having said, “We have a shot of being in production for version one of Optimus hopefully next year.”
That’s next year, as in 2024. Within 12 months.
Not surprisingly, the haters came out of the woodwork.
Tesla and Musk are famous for not having a public relations department that throws money around on advertising with the mass media.
That’s why Tesla, and Musk, often get so much negative press.
Here were a few favorites that came out after the Optimus release:
“but A.I. Experts have their doubts” (when referencing Musk’s plan to start production next year) — CNBC
“Is it the ‘friend’ Elon Musk says will cost ‘much less than $20,000’ and allow for ‘a future where there is no poverty?’ Probably not.” — The Verge
“[…] it’s best to take Tesla’s claims with a grain of salt until they are independently verified in practical, real-world demonstrations.” — Ars Technica
Let’s get this right…
When people virtue signal and talk about a world without poverty, and they’re doing nothing at all about it, they are applauded.
And when Musk commits to a world without poverty and reduced carbon emissions — and has already built a technological solution to enable that — it’s discredited as unlikely.
But again, who cares what “they” say?
If we care about the human race, we don’t want Musk and his team pandering to the passive aggressive detractors. We want them building!
I’m shaking my head again…
Can we think of any other technologist that has a better track record of actually doing remarkable things and achieving radical breakthroughs with technology?
Also, let’s consider this…
Below is a video from August 2021, when Musk and his team first announced Optimus.
The presentation is about 7 minutes long for those that might want to watch it. You can see it here.
The video is obviously a human dressed up in a robot suit. It was completely meant to be in good humor. But Musk was laughed at by the robotics and artificial intelligence “experts,” as well as by the media.
They’re not laughing now.
Consider how incredible Tesla’s progress has been…
In just 28 months, Tesla has gone from an idea and a human-in-a robot suit… to a fully functional, pre-commercial humanoid robot...
Just like Musk said they would. And in fact just like I predicted.
Here’s what I wrote to my readers of my former daily, The Bleeding Edge, at the time…
“We shouldn’t be distracted by all the humor. This is no joke — Tesla will deliver. In fact, I expect to see a prototype as early as next year.”
— Jeff Brown, The Bleeding Edge, September 13, 2021
And that’s exactly what happened.
An early prototype — Bumblebee — was released in September of 2022...
Optimus “Gen 1” was released in March of this year...
And now, Musk is saying that Optimus can be in commercial production within 12 months.
I know who I’m going to bet on. Musk and his robotics team will have a commercial version of Optimus ready by the end of next year. And we should expect to see early production of Optimus within that time frame.
But let’s back up a moment, because I don’t want an important point about Tesla’s incredible progress to be missed.
Let’s gain even better perspective...
Agility Robotics, a company that I like, has been at this since 2015. It too has made great progress. It now has a bipedal robot — Digit — designed for one application: logistics operations.
Or how about Boston Dynamics, a company that has been around since 1992? It has changed hands several times from Alphabet in 2013, to Softbank in 2017, and now owned by Hyundai since 2021.
Above is Atlas, which has demonstrated impressive agility and acrobatics in a controlled setting.
This is purely a prototype, as Boston Dynamics is primarily focused on its dog-like Spot robot, and its Stretch robot for logistics operations, capable of moving boxes and pallets around.
So how has Tesla done it?
How has it completely leapfrogged some of the most advanced robotics companies — who have been at it for years, decades — in just a matter of 28 months?
Tesla is head and shoulders ahead of Agility and Boston Dynamics in robotics right now.
Musk clearly has a team of remarkable engineers and computer scientists that have brought Optimus to life in a remarkably short period of time.
The secret to its success has been leveraging its core competencies.
My long-time readers might remember me explaining how this works in an issue of my past research newsletter, The Near Future Report. Here’s what I wrote:
“When I think about Tesla, I see one of the leading AI companies in the world. Not only does Tesla have the most advanced AI software, it has access and control of its AI network… And that network is its fleet of cars driven by its customers around the world.
This is the incredible part that most don’t see or understand. Tesla designed a software/AI platform designed to operate cars (and trucks). And every one of Tesla’s cars is a “connected” part of its own network.
That’s right, each Tesla car is connected to Tesla via a wireless network, and its “fleet” of self-driving cars is collecting driving data, images, performance data… everything that an AI needs to learn and get “smarter”… every single day. Real-world data is the single most valuable asset necessary to train an AI, and no other company on the planet has the network in place, nor the data, to compete with Tesla."
— Jeff Brown, The Near Future Report, November 2018
This is what most people failed to see for years about Tesla, hard as I tried to make it known…
"Wall Street focuses on demand for electric vehicles. Or how many Tesla's get shipped each quarter. This shows me they still don't understand what Tesla is.
It is NOT a traditional car manufacturing company. It is actually one of the world's most advanced AI companies. That's the big secret most investors haven't figured out yet."
— Jeff Brown, The Near Future Report, January 2020
If you were one of the readers who listened, you bought in at the beginning of November 2018 at a split-adjusted price of $23 per share.
Tesla is now trading at $253 per share — up almost 1,100%. (At its peak in October 2021, Tesla was up more than 1,700%.)
Now, while I’ve long said that Tesla is the largest, most successful AI company in the world, we must remember that it is also — clearly — a manufacturing company, too.
In fact, it is one of the most advanced manufacturing companies in the world.
And now it’s leveraged its skill in building hardware… and combined that with its prowess in artificial intelligence…
And with that, Tesla was able to leverage its full, self-driving software from its electric vehicles (EVs)… in order to give Optimus enough intelligence to perform a wide range of tasks, including those that require fine motor skills required to manipulate a fragile object like an egg.
I have a Tesla, and the full self-driving software package. I got one because I wanted to work with the AI and test it.
Despite what the detractors say, its incredible. While it continues to get better and better, it is already at the stage where it can often take me point to point without any driving required at all.
This is why Tesla has an inherent advantage over all the other robotics companies.
It has already developed fully autonomous software for machines — its EVs — that are in production and in use every day.
Billions of miles have been driven on Tesla’s autopilot. That data has been used to radically improve Tesla’s AI at a pace that no other company, or government, can match.
All of this technology is at the outer limits of artificial intelligence and directly applicable to robotics.
So despite what we might read in the press, it’s happening, right now.
Which is why Asimov’s laws of robotics, or some variation thereof, are critically important right now.
If only Asimov were still around to have seen it.
