Elon Musk's Big Bets: Starship & Tesla's AI Fleet

TL;DR

SpaceX just caught a rocket with giant chopsticks, a massive engineering win. Meanwhile, Tesla's AI demo of its Optimus robot might not be what it seems. #VentureStep #SpaceX #TeslaAI

INTRODUCTION

What did you get done this week? ³It’s a simple question, but one that cuts to the core of productivity—and it’s a question Elon Musk famously posed to the former CEO of Twitter before deciding to buy the company. ⁴⁴⁴⁴ It perfectly encapsulates a mindset of relentless execution, a theme that defines two of his most ambitious companies, SpaceX and Tesla. Recently, both companies showcased staggering progress that pushes the boundaries of engineering and artificial intelligence.

In this episode, host Dalton Anderson dives into the monumental achievements and ambitious plans unveiled by both SpaceX and Tesla. We start by looking to the stars, analyzing SpaceX's fifth test flight of Starship, which featured a groundbreaking landing method that could redefine space travel. 5 This wasn't just another launch; it was a successful demonstration of a system designed to make humanity a multi-planetary species.

Then, we shift our focus back to Earth and explore Tesla's "WeRobot" event, where the company rolled out its vision for an autonomous future. ⁶From the RoboTaxi and RoboVan to the latest on Full Self-Driving and the Optimus 2 robot, Tesla presented a future driven by AI. ⁷However, Dalton also raises critical questions about the presentation, particularly concerning the Optimus demo, and explores whether the showcase was a genuine leap forward or a cleverly disguised illusion. ⁸

KEY TAKEAWAYS

• SpaceX’s “chopstick” landing system, Mechazilla, is a revolutionary step toward rapid reusability by eliminating heavy landing legs, increasing payload, and drastically reducing turnaround time. ^9999999
• Tesla’s vision for autonomous transport relies on drastically low cost-per-mile projections—20 cents for the RoboTaxi and 5-10 cents for the RoboVan—but faces significant infrastructure and regulatory hurdles before becoming a reality. ^{10101010101010}
• The path to full autonomy highlights two different strategies: Tesla is betting on a scalable, camera-based system, while competitors like Waymo use more expensive LiDAR technology that is already operational in several cities. ^1111111111
• Tesla’s Optimus robot demonstration may have been misleading, as multiple reports from attendees suggest the robots were being remotely controlled by humans, raising questions about transparency in AI development. ^{12121212121212}

FULL CONVERSATION

Dalton: Welcome to Venture Step Podcasts where we discuss entrepreneurship, industry trends, and the occasional book review. ¹³ I would like to start this episode off with a question. The question is, what did you get done this week? ¹⁴ It's a question you should frequently ask yourself when re-evaluating your productivity throughout the week. And it's also a question that Elon Musk asked Parag Agrawal, who is the former CEO of Twitter. ¹⁵

Dalton: Parag reached out to Elon Musk and asked him to reduce his frequency of discussing Twitter's downfall and to join the board to help lead the company back in the right direction. ¹⁶

Dalton: Elon Musk then replied, "What did you get done this week? This is a waste of time. I'm not joining the board. We'll make an offer to take Twitter private." ¹⁷

Dalton: We all know how that went. Twitter is no longer Twitter; it's X. 18 Elon gets a lot of stuff done, and that's the whole point of this question. It's a meme that has resurfaced in recent weeks because of all the stuff that Elon has gotten done, and that's what we're going to be discussing today. 19 We're going to be discussing SpaceX and Tesla. SpaceX had a successful fifth test flight of their Starship, and then Tesla had their WeRobot event, which included the rollout of the RoboBus and RoboTaxi. 20 Optimus 2 was also demoed... ish. We'll discuss why I say "ish" later on in the episode. 21

Why SpaceX is Trying to Colonize Mars

Dalton: First on the list, we're going to discuss SpaceX. ²²SpaceX is a company that has really pushed the boundaries of what engineering can be and what's possible. ²³Before SpaceX, launching rockets into space was incredibly expensive, costing taxpayers around $200 million per launch. ²⁴The primary goal of SpaceX is to make the human race multi-planetary, and the first step is to colonize Mars. ²⁵To accomplish that, there needs to be significant optimization with price and technology. ²⁶

Dalton: The first thing SpaceX did was make rockets reusable. ²⁷SpaceX was the first company to make a reusable rocket that flies out of orbit and lands back down, which they achieved with their Falcon 9 rocket. ²⁸By making just stage one reusable, you can save about 30%, which made SpaceX substantially cheaper than other vendors for shipping things to space. ²⁹This competitive advantage is huge, even as other companies like Blue Origin work to catch up. ^30303030

The Problem with Landing Legs

Dalton: SpaceX is still pushing the boundaries of what is possible. ³¹Recently, they had their fifth flight of Starship, but instead of landing with legs, they used a tower with arms, known as Mechazilla or the "chopsticks," to catch the ship. ³²The reason for this is that building legs strong enough to withstand the weight of the massive Starship adds a lot of mass to the rocket, which in turn reduces the payload. ³³If you reduce the payload, you reduce the potential revenue. ³⁴

Dalton: With the Falcon 9, some customers were paying SpaceX to not have legs on the rocket. ³⁵

Dalton: The Starship is way bigger than the Falcon 9, so its legs had to be much beefier, further decreasing the revenue SpaceX could get. ³⁶ There were other issues as well. The legs on Starship have collapsed before because the ship is so heavy. ³⁷Additionally, when the ship lands, the superheated plasma from the engines melts steel, makes concrete explode, and can damage the ship itself and the landing pad. ^38383838Finally, the process of attaching and detaching the legs is manual and time-consuming, which slows down the turnaround time for launches. ^39393939

The 'Chopstick' Solution: How Mechazilla Works

Dalton: All of those issues combined led to the brainchild of Mechazilla. ⁴⁰ Mechazilla is this large tower with arms on it. Instead of landing the Starship on the ground with legs, they decided to just catch it on these arms. ⁴¹ It's crazy how tight the margins are. I originally thought the chopsticks had a wider margin of error, but they are actually small little nubs that come out of the ship for the arms to catch. ^42424242

Dalton: You're like, "Hold on, hold on. Let me land it on these little nubs while this thing is closing so we can launch faster." ⁴³

Dalton: The ship recalibrates and straightens out at the last second, and while that's happening, the arms are closing at the exact same time. ⁴⁴It's hard to understand how complicated that really is. ⁴⁵They've overcome a big issue here, and the main goal is to have a faster launch. ^46464646

What's Next for SpaceX? Orbital Refueling

Dalton: The next big thing SpaceX is working on is orbital refueling. ^47474747To execute longer missions, like a trip to Mars, they need a way of refueling the Starship in space with another Starship. ^48484848They plan to do this using a "belly button" technology they designed, but it comes with its own challenges. ^49494949How do you navigate these two massive rocket ships, ensure precise and stable docking in space, and safely transfer supercooled liquids? ⁵⁰

Dalton: Looking forward, they are also working on a lunar lander for NASA and want to substantially increase the frequency of launches. ^51515151The goal is to manufacture a Starship every 24 hours and have multiple launches per day, with turnaround times of just a few hours instead of 20 days. ^52525252The chopsticks are key to making that possible. ⁵³

Tesla's WeRobot Event: A Glimpse Into the Future

Dalton: Now, let's talk about Tesla. ⁵⁴Tesla recently had its WeRobot event, which covered new features and products they're releasing. ⁵⁵The first thing that encapsulates the whole event is that Full Self-Driving is entering a wider beta. ⁵⁶The new release is going to have a threefold increase in capabilities, and a future update will have a sixfold increase on top of that. ⁵⁷

The Promise and Challenge of Full Self-Driving

Dalton: There are still issues regarding accidents and regulatory questions on how to mandate and monitor Full Self-Driving. ⁵⁸ Another question is what's going on with Tesla compared to competitors. Waymo, a subsidiary of Google, has been doing full self-driving ride-sharing for a bit now in many major cities. ⁵⁹Tesla was first to the market and has been talking about this for the longest time, but their approach is different. ⁶⁰

Dalton: Tesla uses cameras, which are substantially cheaper than the LiDAR that Waymo uses. ⁶¹Full self-driving is possible with LiDAR, but it hasn't been done with just cameras. ⁶²If you can figure out cameras, you can definitely do LiDAR later, but I don't think you can just transfer LiDAR information over to cameras. ⁶³My issue with Tesla is, how long are you going to talk about Full Self-Driving before you roll it out? ⁶⁴They are still holding out this treat while also showcasing the RoboTaxi and RoboBus, which they say are coming in 2026 or 2027. ⁶⁵

RoboTaxi and RoboVan: The Economics of Autonomy

Dalton: They're saying the cost per mile for the RoboTaxi is around 20 cents, and for the RoboVan, it's five to 10 cents. ^66666666

Dalton: How crazy are those numbers? Those numbers are insane when you consider AAA estimates 65 cents a mile for a private vehicle, and taxis are $2.50 to $4.00 a mile. ⁶⁷

Dalton: The RoboVan can carry up to 20 people and has a modular inside where you can change the configuration of the seats. ^68686868It looks a lot like the bus from the movie

iRobot. ^69696969 However, for these vehicles to roll out, a few things need to happen. One, they need Full Self-Driving approved. ⁷⁰Two, they need to set up a wireless charging network, as these vehicles don't have charging ports. ^71717171This requires a massive new infrastructure for both charging and power grids. ^72727272

The Optimus Robot: Impressive Tech or Misleading Demo?

Dalton: The last item is the Optimus robot. ⁷³The Optimus robot has human-like abilities, with a lot of dexterity in its fingers to pick up fragile objects like eggs. ^74747474The engineering on the hands is crazy. ⁷⁵ But when the Optimus was being showcased, there's some confusion. Those robots were reportedly manually controlled. The people controlling the robots were also speaking as the robot. ⁷⁶

Dalton: Several sources, including attendees and analysts, noted that the robots changed their voice and mannerisms throughout the event. ⁷⁷You could talk to the same robot, and its voice would be completely different later on. ⁷⁸The level of detail in the conversation and the responses just seemed like it was another person. ⁷⁹Neither Elon nor Tesla have commented on this. ⁸⁰

Dalton: I just think it's misleading... when you demo like "we are AI" and "we're rolling out fleets of autonomous vehicles soon," and then people find out it’s being tele-operated by a human. ^81818181

Dalton: I do respect companies that do live demos. ⁸²I appreciated Google's live demo and Meta's live demo; they had mistakes because it's live. ⁸³If you can't demo it live, then you shouldn't do a whole showcase thing. ⁸⁴If you can't do it live, then don't do it. ⁸⁵

RESOURCES MENTIONED

• SpaceX
• Tesla
• Twitter (X)
• Elon Musk
• Parag Agrawal
• Starship
• Falcon 9
• Blue Origin
• NASA
• WeRobot event
• Optimus 2 robot
• RoboTaxi
• RoboVan
• Waymo
• Google
• Uber
• Lyft
• iRobot (movie)
• Grok AI
• Nvidia
• Jensen Huang

INDEX OF CONCEPTS

Elon Musk, Parag Agrawal, SpaceX, Starship, Falcon 9, Mechazilla, chopsticks, reusable rockets, payload, orbital refueling, belly button technology, Mars colonization, NASA, Blue Origin, Tesla, WeRobot, Full Self-Driving, FSD, RoboTaxi, RoboVan, Optimus 2, Waymo, Google, Uber, Lyft, iRobot, LiDAR, cameras, wireless charging, Grok AI, Nvidia, Jensen Huang