Pezpunk
Active Member
So yeah exactly like I said. Zero evidence. Just wishful thinking.
-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Tesla, TSLA & the Investment World: the Perpetual Investors' Roundtable
- Thread starter AudubonB
- Start date
TheTalkingMule
Distributed Energy Enthusiast
Well.......except Elon saying it's gonna happen. I don't care one lock about this bot, it's clearly just to send to Mars, but Elon saying something is coming carries a certain weight with me.
dhanson865
Well-Known Member
It won't rip off his arms because they designed it to be mechanically weak in many ways. But it could grab his hand and not let go. So worst case is it grabs his hand, doesn't let go, someone has to come shut it down and reboot it or use a manual release.
I agree with this, AS LONG AS Tesla can successfully make Optimus as they envision it. If successful this could be one of the greatest inventions of all time and the largest boon to TSLA in the company's history, BUT there is always the chance their vision does not succeed too. Victory is not guaranteed, although I do believe they will pull it off.
I think we'll have a much better comprehension of the product after AI Day 2, particularly to it's development progress.
I almost gave you a disagree for no other reason than you missed an opportunity to quote Monty Python.
My only input in the discussion.
Making Optimus functional and productive is exponentially easier than FSD. The dynamics at play are nothing alike. Making Optimus a highly efficient and more production factory line worker than human is not even in the same class as getting FSD is 99.9999%
I actually expect Optimus progress, productiveness, and abilities to leapfrog FSD over the next one to two years. I’m actually surprised more people don’t get this
Agree on the no published specs part, as for "seen zero evidence anything approaching a working prototype exists" also having any significance, I'll just leave this here...
philw1776
Member
Very skeptical about Optimus progress. Disagree it's easier than still non-functional FSD. Road environment and rules much more constrained than the rest of physical reality. Agree that easiest Optimus application would be a well constrained environment like a Tesla factory floor, but even then...
willow_hiller
Well-Known Member
I wouldn't worry too much about the current cultural relevance of the Daily Show. Their viewership has been dropping for years. They're pretty much at the bottom of the Late Night ratings, with about 400k viewers:
I watched the segment myself. It will resonate with uninformed people that hate Elon for other reasons, but not much else. Shameful of them to use clips from O'Dowd.
Last edited:
Road environments are absolutely not more constrained.
Just an example. For Optimus to be a productive and efficient factory worker, there’s only an X number of machines it needs to learn to function. It then only needs to be able to learn the factory floor environment, which does not change very often. It’s working in a relatively static environment with machinery that doesn’t change often
That dynamic doesn’t even remotely compare to the random elements that FSD could and will encounter on a single drive. Not even close
Gigapress
Trying to be less wrong
Imagine the CEO and Chief Product Architect of a company publicly announces something like “So...remember that big highly-publicized engineering recruiting presentation and Q&A we held last year in which we announced a new product that we had been secretly working on with all our best engineers and then at the end of the presentation explicitly invited exceptional engineers around the world to apply to help on the project? Remember how we said it had become our #1 top priority for research and development? Remember how we announced shortly after the recruiting event that it was a smashing success and the job application rate had increased 100-fold? Well guess what: we’re going to do a sequel recruitment event in a few months, we’ll give a sneak preview of the product with cool and exciting updates, and hopefully we will be able to demonstrate a working prototype by then!”
Imagine also that this company and its leader have an extensive track record of announcing technology goals that sound impossible and the later achieving the targets, as well as an extensive track record of attracting incredible engineers and giving them an amazing work environment for ambitious nerds.
Does that sound like a situation where the product development is actually still in the pure fantasy stage, that the prototype is not even approaching functionality, and the whole thing is nothing more than wishful thinking?
Furthermore, does that give you the right to insult and belittle somebody who doesn’t believe that and who, as a rare contributor to this community, might be nervous about sharing their thoughts?
Last edited:
I have thought about the "machine friendly world" , it's parameters and consequences, for some time. You are right, Optimus will initially exist in a tailored environment performing specific tasks that currently require human agility and senses. It will excel right out of the gate at it's intended function.My only input in the discussion.
Making Optimus functional and productive is exponentially easier than FSD. The dynamics at play are nothing alike. Making Optimus a highly efficient and more production factory line worker than human is not even in the same class as getting FSD is 99.9999%
I actually expect Optimus progress, productiveness, and abilities to leapfrog FSD over the next one to two years. I’m actually surprised more people don’t get this
Optimus that will walk the dog, pickup tonight's groceries, make dinner and burp the baby is a long way down the road from now. Know the difference!
MartinAustin
Active Member
Exactly. At the time they had a human dancer dressed up in an Optimus outfit, doing a little demo/dance routine... who was under the impression they were just then sitting down to make Optimus? It's more probable they had already been working on it. Just nothing show-worthy. One year is a long time in Tesla Land, and there are no rules, no accepted practices for humanoid robot production - and the prototype team are probably quite motivated.
MartinAustin
Active Member
Just watched "Autoline Daily." According to the NHTSA, the current death rate on the roads is 1.25 per 100 million miles driven. (7% higher than last year)
It so happens that FSD Beta has had about 100 million miles driven. With zero deaths, and I'm willing to guess zero injuries, zero serious crashes and perhaps zero minor crashes. (Yes there have been a lot of scrapes and kerbed alloys)
It so happens that FSD Beta has had about 100 million miles driven. With zero deaths, and I'm willing to guess zero injuries, zero serious crashes and perhaps zero minor crashes. (Yes there have been a lot of scrapes and kerbed alloys)
dedicatedtek
Member
I like your thinking, but I don’t think this is the reason
Tesla isn’t worried about Osborning their current deliveries. Demand is just too strong at the moment. Maybe a handful of die hards will delay, but can’t see a ton of that, particularly when they have Hertz and Autonomy out there sucking up all the cars they can get their hands on. If Tesla wanted to dump a bunch of cars with HW3, Hertz would gobble them up.
I would add some further caution, that the Tesla-bot will most likely have a similar release schedule/timeline as FSD.
As a reminder: FSD had not only working prototype hardware, but also consumer-product release hardware since 2016, that is 6 years!
Yet, the software is still in restricted participation beta. I expect the Tesla-bot to be also released in hardware first with rudimentary software which then gets updated many times regularly as the NN is trained with more and more collected data from the hardware operating. Also note, that FSD had 3+ hardware revisions since that original release in addition to the many many software revisions.
Similar threads
- Locked
- Locked
