There were two Teslabot videos. The first has a discussion with James Douma. James describes his perspective of the advances in neural nets. He described how GPT-3 created a foundational capability by cracking language. He believes the Teslabot will leverage neural nets to crack robotic methods for bipedal movement and mastering identifying and picking up objects.
There was a panel discussion of Teslabot by several Tesla youtubers.
Bradford Ferguson, Scott Walter, John AKA Dr Know It All, and Farzad Mesbahi had a panel discussing the Tesla Bot’s impact on society, job creation, and more. Elon Musk and Tesla are set to showcase the Bot at their AI Day 2 on September 30th, 2022.
The general consensus was that Teslabot will be a job creator.
Brian Wang is a Futurist Thought Leader and a popular Science blogger with 1 million readers per month. His blog Nextbigfuture.com is ranked #1 Science News Blog. It covers many disruptive technology and trends including Space, Robotics, Artificial Intelligence, Medicine, Anti-aging Biotechnology, and Nanotechnology.
Known for identifying cutting edge technologies, he is currently a Co-Founder of a startup and fundraiser for high potential early-stage companies. He is the Head of Research for Allocations for deep technology investments and an Angel Investor at Space Angels.
A frequent speaker at corporations, he has been a TEDx speaker, a Singularity University speaker and guest at numerous interviews for radio and podcasts. He is open to public speaking and advising engagements.
I’m still calling BS. This will be delayed and delayed just like so many other Tesla new techs a la Cyber truck because there is no current tech that can fit in this form factor (especially usable battery life configs that last more than 30 minutes. I’ll be happy to be proven wrong.
True
I suspect half the economy is low hanging fruit for Teslabot, jobs that require little more than solving machine vision and having a humanoid robot that can be a platform for neural network task training in Dojo. Driving of course is a skill it inherits from it’s parents. With some retraining that means Teslabot can turn a vast range of existing work vehicles into autonomous robotic systems by sitting down and driving them. This is everything from Forklifts and Tractors to Zamboni . Even earlier there are an endless array of simple ad hoc factory tasks that are difficult/expensive to automate with dedicated machinery. Often as simple as picking something up and putting it in a box or into a different machine. Of course there’s the last few feet problem of robotic delivery trucks, loading the trucks and then carrying packages to a doorstep or delivery point. Teslabot can also make conventional delivery vehicles driverless.
There would be an enormous amount of training but that’s why Tesla would iteratively grow Dojo. There would be a lot of jobs for people to teleoperate Optimus robot bodies on tasks they were being trained for – both turning physical tasks into things that could be done remotely from anywhere and over time training NN to do them without a human – much as Tesla did with tens of thousands of human drivers.
“With some retraining that means Teslabot can turn a vast range of existing work vehicles into autonomous robotic systems by sitting down and driving them.”
Teslabot has blindspots, self driving cars are covered in cameras in different angles to eliminate blind spots. If you add a little camera conversion kit to the vehicle it could work though.
I doubt there IS a lot of ‘low hanging fruit’ for Teslabot – everything you would want a humanoid robot to do (other than ‘be a novelty’) will be require a fairly high level of physical intelligence – or we’d have already programmed regular robots to do the task. In fact I suspect many of the tasks that human workers do might be better done with a somewhat less humanoid robot. E.g. add a ‘camera finger’ to the robot’s hand so it can work in narrow spaces that are awkward for a human. Or tiny hands for tiny parts Or really big, flexible hands for carrying furniture.
“In fact I suspect many of the tasks that human workers do might be better done with a somewhat less humanoid robot. ”
There are already tools that humans use to do things like carry furniture, so all you need to do is train the Teslabot to use them. Why reinvent the wheel?
Moving furniture is NOT low hanging fruit for a robot. A Teslabot might easily be able to pull a chair out from the dinner table – but not move a dinner table from a truck into a house. Even using a dolly or hand truck to handle the weight, that is going to be a complex task. Let alone training or programming it to handle arbitrary pieces of bulky furniture the way a human mover can.
“Fetch me a beer.” “Take this bag to the kitchen, put it on the floor and come back.” Those are the sorts of things that may be ‘low hanging fruit’.
Even “Carry this bowl of soup to the table and set it down” may be challenging for a Teslabot (i.e. sloshing liquids).