1,000 Tesla bots in 18-MONTHS - the roadmap to Optimus w/ Scott Walter, PhD

#optimus #TESLABOTS #robots

[davidmarkmann]

Another great video. If you are robot smitten then a 6-12 posting to Nevada to set-up the Optimus line will be a dream gig.

[budgetaudiophilelife-long]

🙋‍♂️THANKS BRIAN, 🤗 FOR HOSTING SCOTT 🤔…SHARING VERY INTERESTING NEWS 👍💚💚💚

[rainflowers]

Interesting conversation, it's going to be fascinating to see the progress of Optimus over the next year or two.

[JDLuke]

I legit laughed at Keebler. Guess that puts me in the demographic.

[dennisfoster]

Last week I saw a bot on display at the Tesla showroom in Northpark mall in Dallas.

[BongoWongoOG]

Scott, priceless commentary as always. Brian, entertaining and knowledgeable. Game changer for me is if/when bots teach themselves by watching and mimicking video, then use repetition to improve the learned function/activity. That what humans basically do. Next 12-24 months is going to be very very interesting times.

[bru]

18:00 Scott points out the most fundamental advantage that Tesla has for building bots. They already an internal business case. No customer, no sales, no marketing is required to get this product off the ground years, if not decades ahead of anyone else.

[markorcutt]

Titanium has limits to solvents to prevent I believe cracking at the surface due to chlorine contamination. The machining coolant has to be specialized also. I remember having a pile of chips on a mill catch fire when not enough coolant was being applied. Elevated temperatures in an oxygenated atmosphere create problems beyond just a discoloration. I also remember reworking Titanium welds. The cracks that looked like little pores could travel deep into the material, required multiple inspections with fluorescent penetrant to verify the removal was completed.

[skinnymoonbob]

Exactly, lease will be the right business model. A SaaS revenue model with optional apps that enable functionality.
2 years at the unveiling of Optimus I thought it would take 10 years at least to scale. It seems to become real in 2 more years, at least the start of simple factory work.

[colinkeizer]

We'll have access to a LOT more titanium, soon, from the recently announced deposits in southwest Norway. Vanadium, too, as well as phosphate. I'm really fascinated to see if we find titanium 3D printing suddenly become more popular, especially in Europe.

[jasperkong]

Thank you Brian. You're absolutely fabulous 👍🏼

[edhill]

Think of the Semi truck. A very slow ramp up as they test the truck. Optimus will take the same approach so thinking there will be 1, 000 BOTS in 18 months is likely far too optimistic.

[h.e.c.]

Great points from both of you, thanks for thorough analysis. One point I don't think either of you mentioned yet in either of videos is the "critical mass" (amount) of Tesla bots in use to to speed up their learning significantly and what kind of correlation we can expect between amount of bots in use, or better say their "mileage" (operational hours), and rate of their improvements. I look forward to next videos about Tesla bot from both of you, perhaps you can even invite Dr Know it All.

[williampmcd]

Bots could economically mine landfills for materials, conceivably reclaiming those areas for other use.

[kevtheobald]

I agree with the robot brain being the biggest challenge.

I also believe large scale production will be harder than expected just because no one has done a large robot with this complexity at large scale. With that said, Tesla's production experience and willingness to try new or out of the box, unboxed 😊, methods makes me believe they will find a method for cranking out tens of thousands a week of the Teslabots. When is the question, not will they.

I do wonder if they would spin off just a set of arms to do stationary type of work. In Modesto, California there is one of those robotic coffee shops where a pair of robotic arms is hung above a work station making coffee. I know there are companies in the robotic kitchen space, but I doubt they could make robotics as cheap as Tesla and if their system as amazing AI as well, I am sure the hundreds of thousands of fast food/beverage businesses would be onboard with just buying the robotic arms. No battery to worry about, they can have a dedicated space where they do their business that is probably smaller than what humans need and the usual benefits of not needing meal breaks, medical insurance, etc.

Imagine a pharmacy that just needs one pharmacist but can fill hundreds of orders a day because the Teslabots arms would fill the orders. The pharmacist just has to consult the customers. How much money would Walgreens save cutting their pharmacy staff by 75%?

[haifenghe]

"Tesla is their own customer", well said. Same goes with SpaceX: Starlink is SpaceX's customer. Amazon is their (AWS's) owner customer. Google is Deepmind's customer. And countless many others are playing the same in other industries, sometimes different companies, but same owner.

[richyt]

I wonder what the estimated downtime is for the bots, esp initially. How reliable are the actuators? How are the moving parts lubricated? What wears out first?

[Crenor]

I would think more like computer gear. (servers, big backend items) upfront cost of bot - with standard 20% ish margins and a 25% of that cost total as a per year maintenance cost that covers updates and warranty of everything.

[ianhoward]

My own model has internal sales for the first 5000 and will eventually build themselves.
Build cost of $20k.
Also $24k annual lease to others.
Run rate ramping to 20m by 2035, with cumulative leasing of over 150m
150m*$22k profit = 3.3t pa 🤯

[ChuckCorcetti]

Don’t have a Twitter account don’t want one however, there is a robot in Scottsdale at the Tesla Scottsdale Fashion Center showroom downtown.