Blue highway
Well-Known Member
- First Name
- Steve
- Joined
- Oct 15, 2021
- Threads
- 5
- Messages
- 2,700
- Reaction score
- 4,249
- Location
- Oregon
- Vehicles
- Mach E Premium SR RWD
A few things are true at the same time...Look at how difficult it has been to get autonomous vehicle software working, and it only has to do ONE TASK (stay between these two lines). Admittedly is also has solve for edge conditions, but edge conditions still based off accomplishing only the single task (stay between these two lines). Additionally, this software operates in a piece of hardware (vehicle) that only operates in a 2-D environment (go forward/back; go left/right) and is inherently "stable" (turn car off, car stays in place even if power is removed).
Contrast that effort now, and consider the effort required in order to get autonomous software working for humanoid robots. Not only do they need to do many tasks, they need to LEARN how to do many tasks (e.g., need to be able to handle situations never before seen) in real time. Think again they also need to do these tasks while handling edge conditions, but not just edge conditions against a single task but edge conditions against each task. They now need to operate in a 3-D environment (go forward/back; go left/right; go up/down). They are inherently "unstable" (turn robot off or power is removed, it falls over) so they need to be designed to be stable or solve for stability.
These are just a few differences. But, I have what I think is a more profound personal opinion and that is, why do they need to be humanoid? There is no universal law that says they need to be. The argument is we live in a world created for humans, but why can't a robot have 4 legs or 4 arms or wheels or any other design and still conform to human infrastructure? I think the humanoid form, for now, is misguided.
I'm also thinking economically. I think about Tesla selling cars, and I think how many of them they sell per year, each costing tens of thousands of dollars. Now I think about robots each costing tens of thousands of dollars. People buy cars because they NEED them, and if we are honest nobody today really NEEDs a humanoid robot. Even though people NEED cars, Telsa after all these years of predictions is still not the largest automaker by volume. Not by a longshot (last I checked I think they were something like number 16).
So what really is the market for a Tesla robot that costs nearly the same as a car? Do we really think that soccer moms already deep in debt on their 72-month car notes going to also take out a $750 / month note on a humanoid robot? It's difficult for me to really think that will be the case. I'd love for anyone who disagrees to maybe show some math how a robot the cost of a car will somehow be any more profitable or marketable than their existing car market? Am I missing something?
1) a humanoid robot is cool in a sideshow sense... just look at CES this year... there were dozens of the things doing "useful" things like dancing
They drew huge crowds.2) when you spend about half your life in a world where nobody tells you "no" and you have state level resources at your disposal, you pursue what interests you... i.e. cybertruck, robots serving drinks, moving mars, etc... This is not about a strong business case.
3) I save a place in my thinking for things I didn't know I wanted till long after seeing them. I had a "smart" phone as prototype around 2002... way back when Palm Pilots were a thing... It was useless and held no appeal to me. But like most folks, I now think a smart phone is indispensable. So.... maybe I just haven't figured this out yet. Maybe my second car should be a robot? Or maybe humanoid robots are still the sci-fi fantasy they have been for the last 100 years...
4) The first time I heard of an electric car... the tesla roadster or the leaf... I had less than zero interest... now, I expect to never buy an ICE car again.
Sponsored