Tesla AI

FSD, Robotaxi, Optimus, Flying Car, predictions

Trying to buy a robotaxi at 21

In 2019, I was watching YouTube videos of Elon Musk and TeslaTubers during my business school classes. My family had set money aside for me to go to college, god bless them. I found business school classes boring, and I finished a year early, so there would be money left over for me to launch my life.

Upon graduation, my plan was to build a startup, write a blog, grow my e-commerce shop, invest in stocks, and buy a Tesla which I could rent out to cover the payments. 

The startup dissipated immediately upon graduation. 

I didn’t know about Substack, so I started a blog on my personal website. In one of my three posts, I predicted that Tesla would be worth $500 billion in 2024, a 15X increase from 2019. That prediction came true.

The summer after college, I decided to do two things: invest everything I had into Tesla, and buy a Tesla. I flew to Oregon, where my cousin lives, and got an Oregon driver's license. There, I could buy a Tesla, tax-free, and get a cash rebate. 

I was thinking that over the following 18-24 months, Tesla would incrementally update the AI software making my new car into a self-driving vehicle, thus appreciating the vehicle beyond what it cost me to buy it brand new.

My family thought I was retarded.

I was sitting outside a sushi restaurant in August 2019, telling my mom and sister about the plan to go to Oregon and get a Tesla. 

“What does dad think?” My sister said.

“He thinks I’m crazy,” I said. I thought about it. “Yeah, I’m crazy.” 

She laughed.

I borrowed $50k from multiple lenders and pulled the trigger. I took delivery of a Model 3 at age 21. It was a surreal moment.

I let my aunt do a test drive. She activated Autopilot and the car accelerated up to the speed limit, right before we hit a speed bump! She hit the brake just in time. 

“That wasn’t a good time to activate it,” I said. Mildly embarressed.

In 2020, my Model 3 got a software update to stop at stop signs and traffic lights while on Autopilot. It was cool, but far from the robotaxi holy grail I was hoping for. 

Despite slower than expected progress in the AI department, my Tesla investment grew 10X because of low interest rates, booming sales of Model 3 and Y, and accelerating earnings growth. The full Tesla investment story is the first chapter of my book, which I partially wrote in iPhone notes as my Model Y drove me to the mountains to go hiking.

I was browsing the Tesla website, checking in on my largest holding, and decided to apply for all of the jobs. I got a job at Tesla, working on Tesla Energy Solar Roof projects. While there, I bought a second Tesla, Model Y, getting the Full Self Driving software package for free as an employee. I held onto my Model 3 for another year, hoping it would turn into a robotaxi. It didn’t. Still hasn’t.

Model 3 and Y at my house

I bought the 3 for 54, and sold it two years later for 48. Not bad. 

In winter of 2022-23, Tesla updated my Model Y with “feature complete Full Self Driving”. Unfortunately it sucked at driving. I often had to intervene, and the ride was uncomfortable. 

Things changed in 2024. Tesla shifted their software stack to full neural net. Video in, behavior out. The vehicle fleet grew north of 5 million cars filming the roads. The high quality training data piled up and Tesla upgraded their AI training systems. They built Dojo, a specialty in-house AI training supercomputer.

This past week, my car has been driving around Boston, blowing my mind. It did a U-turn, navigated a busy roundabout, and slowed down for speed bumps. 

My eyes got wide.

A year earlier, it was a nervous 15-year-old with a learner’s permit. Today, it’s a confident 17-year-old. 

This is my car driving me around Boston in September 2024:

Today, five years after buying my first Tesla, there are zero unsupervised Tesla robotaxis on the roads. But that’s about to change. 

A preview from Tesla of their Robotaxi ride-hailing app

Analysis, Predictions, and Musk’s Cock

The future is here, it’s just not evenly distributed. 

In 2025 Tesla will launch and grow its robotaxi ride-hailing app in select cities in America; similar to Waymo, whose robotaxis are currently giving customers rides in SoCal and Austin. But Tesla will scale much quicker because they have millions more vehicles and much more training data. 

A large fleet is needed to win self-driving at scale because a massive amount of high quality, up-to-date data is necessary for the system to learn the infinite edge cases it will encounter, and be able to drive in almost any environment and circumstances. As far as I know, only one company has a fleet of millions of cars collecting data.

Elon’s idea is that, upon regulatory approval, owners will be able to add their vehicle to the Tesla Network whenever they’re not using their car. Ownerse can “make money while they sleep.” I literally bought into this dream five years ago. Maybe this happens. But I think Tesla will own and operate the vast majority of the Tesla Network ride-hailing fleet. 

Andrej Karpathy, former chief of AI at Tesla, said of AI robotics in general: 

“The demo is near, the product is far.”

I agree. It will launch 2025 but it will take Tesla 10 years to scale up the ride hailing service with the majority of growth being in the 2030s. 

I predict that in the late 2020s Tesla stock will go to the moon, again, as the robotaxi network grows, and while SpaceX’s Starship lands on Mars. What a time to be alive. 

I hope Elon doesn’t have to watch from a jail cell, or heaven, as his wildest dreams come true. But when you go against the great state powers, that’s usually what ya get. Just ask Telegram CEO Pavel Durov who was arrested and put in a French jail for not cooperating with censorship “requests”. Will Elon die on the hill of free speech?

I hope Elon gets to see Robotaxi, the Optimus humanoid robot, Starship, and his twelve human children mature into spectacular usefulness and beauty. I wouldn’t bet on it. But also I wouldn’t bet against Elon. 

I wrote about the Optimus robot in Tesla Human Factory.

The Big 10/10 Reveal

On October 10, we will see what’s around the corner for Tesla. The main unveil is their robotaxi which Elon has called “Cybercab”.  I’ve also heard him allude to a few surprises in his back pocket. 

Elon said live on X: “Robotaxi is the main thing we’re going to show. We’re also going to show off a couple of other things.”

Elon is reliable with one thing: It won’t be boring.

At the All-in Summit, in a hilarious exercise of free speech, Elon told the story of pitching an idea to the SNL producers for the week that he hosted the show. 

Elon and friends got on the Zoom call with the SNL people and Elon pitched:

I throw the script down and say, We’re going to find out if SNL is actually live. The way that we’re going to do this is I’m going to take my cock out.

The audience’s jaws dropped. 

If you see my cock, you know it’s true, and if you don’t, it’s been a lie.

I’m gonna reach down into my pants. I stick my hands in my pants…then I pull out a baby rooster. This is my tiny cock. What do you think? Do think it’s a *voice cracks* nice cock? 

Kate’s gonna comes out with a cat…you can see where this is going.

The kids call this Big Dick Energy.

For the 10/10 Tesla event, I predict that Elon pulls out a flying car.

Yes.

That’s right. 

I predict Tesla and SpaceX engineers have collaborated to put boosters on the new Tesla Roadster that will allow it to fly short distance. Just for the hell of it. 

I’m confident in this prediction.

I also predict that an Optimus Bot, perhaps multiple Optimi, will step out of the robotaxi. It, or they, will demonstrate being useful. Perhaps they deliver Elon’s cock.

There is a nonzero chance that we see useful humanoid robots step out of a flying robotaxi.

Investing in Tesla and watching them create awesome stuff has been one of the great joys of my life. I was once all-in TSLA. Today, my portfolio is around two thirds TSLA, 30% High Yield Cash and bonds, and a little bit of AMZN, AAPL and Bitcoin. 

The Fed is dropping interest rates and the AI products are improving. I will likely acquire more TSLA. Every time the Fed drops, TSLA cars become more affordable and Tesla makes more money. At the same time, stocks become more attractive to investors as the retrun on government bonds goes down.

Meanwhile, Tesla keeps shipping updates.

The main point of this is essay is that my car is driving itself. It couldn’t drive when I bought it, and now it can.

Like electricity, telephones, and commercial air travel, robotaxi transportation is a civilizational milestone.

It’s a great time to be a consumer and an investor. 

I looked forward to $4.20 robotaxi rides to the airport, grocery store or across the state. In theory, a robotaxi does go not have to go back to specific place to sleep at night like human drivers do. It has no problem sitting and waiting for its next summon, anywhere, for any duration. A sufficiently advanced system will predict demand and have them eerily ready to go before you press summon.

Head-to-head, an 11-minute wait time for a $30 Uber across town has no chance of competing. What’s more, there’s no random person inside a robotaxi as there is with Uber.

Robotaxi will only operate in cities that allow and help it happen. Cities want this. They compete for talent and tourism. SF, LA and Austin already have driverless robotaxis in operation. It’s not a matter of if, it’s a matter of how fast it scales.

I’d rather invest now, well before it’s obvious to other investors that this is a super profitable enterprise that is actually happening. 

The average person has no idea how far along Tesla AI is. 

The people who own the rights to the profits of the most useful AI systems will be rewarded financially.

Tesla is a basket of tech companies. It’s largely an AI robotics company cranking out bold new products, led by one of the great heroes of the century. The people who have trusted in Musk and Co to mass produce useful products have been rewarded for their patience and commitment, despite the uncertainty, lies, volatility, and tardiness. TSLA has been, and will continue to be, a vehicle for making people rich.

When I was 18, I looked at the macro situation. Industrialization, population growth, globalization, and the internet, had driven massive growth over the past 100 years - all approaching maturity. We can't globalize a second time. There’s almost nowhere left to connect to the internet. I asked myself, What could drive growth to a similar degree for the next decades? The only answer I could think of was AI. That was 8 years ago. It’s starting to come to fruition. 

When I first invested in Tesla in 2016, I was buying a car company. Conveniently for me, it morphed into much more than that. TSLA is today one of the best publicly available AI investments because of their engineering prowess, pace of innovation, and head start in developing the critical ingredients for cash flow positive useful AI: 

• Data 

• Compute 

• Robotics 

• Manufacturing 

Similar to how Amazon built a hyper profitable web services business alongside a massive e-commerce store, Tesla is building an AI services business alongside their massive car business. 

Every current and future Tesla car has a powerful Tesla AI computer in it. When the cars aren’t in use, which is most of the time, that very large distributed compute power can be used for other services. 

By 2030 it will be obvious that thinking of Tesla as a car company is like thinking of Apple as a telephone company.

Tesla robotaxis aren’t a car with a computer in it, they are a computer with wheels on it.

Today, people think of AI and think chatbots and pretty pictures. Tesla is barely in the conversation. But yet, it has the most advanced real-world AI robots in operation. Tesla customers today are paying $99/month to have Tesla AI drive them around.

This is just the beginning. 

The chips, training clusters, sensors, batteries, and actuators Tesla has developed are applicable to other use cases. Humanoid robots and Robotaxis are obvious big ones. But like Amazon Web Services, it’s hard to even fathom how broadly useful this tech can be. AWS runs a THIRD of the internet, from Netflix to government websites.

Tesla has the opportunity to be a major bone in the spine of AI. Much of “The Cloud” for AI could be, and I think it will be, a mix of AI computers on Tesla vehicles, and Tesla compute centers.

The internet took 30+ years to develop. AI could take even longer and be more impactful. Tesla AI is a 40-100 year investment. 

As of Q2 2024, Tesla has built $2.5 billion of AI infrastructure. Today, Tesla is hiring for dozens of positions in “AI & Robotics” for Tesla Bot, Tesla Bot Tooling, Simulation, Dojo [AI training supercomputer], AI evaluation, Self-Driving, Autopilot AI, AI infrastructure, AI platforms, and AI tooling.

Four of Elon’s companies, X, xAI, Neuralink, and Tesla are working together to create the world’s most useful AI. 

The Musk Empire

Rooting for businesses, for me, is more fun than rooting for sports teams. Rooting for Tesla sometimes feels like rooting for the Patriots when they had Tom Brady. In regard to building AI, Elon said, “Play to win, or don’t play at all.” Brady said the same thing.

Getting to the Beginning

Yesterday, 9/28, Tesla shipped ASS - Actually Smart Summon to my car. When I press Summon on the Tesla app, the car will come find me in a complex parking lot.

Since getting my first Tesla, I’ve put many thousands of dollars into tires, insurance, charging, loan interest, and depreciation. I’d be substantially financially richer if I had bought and held more TSLA instead of multiple Teslas. That’s okay. It was fun. I watched the birth of the robotaxi from the driver's seat.

On my Model Y, the cabin cam detects when I close my eyes or look at my phone. It gives me a warning to pay attention. If if don’t oblige, it will beep, come to a stop, and disable FSD. If that happens three times, the system suspends me from using FSD for a week. I get put in timeout every other week. 

Yesterday I went to activate FSD but it was gone. I was like, Dang, I actually wanted it to drive me home. It has earned my trust and I can relax while it drives. 

It’s turned the corner. 

A map of Boston looks like spaghetti. Roundabouts, speed bumps, 5 way intersections, and Masshole drivers. I just let the car drive because I don’t feel like dealing with it. 

This product is currently $99/month or $8k for lifetime. The take-rate and the price will increase as it continues improving and it earns the trust of more drivers. 

The value of a Tesla with lifetime FSD software will rise dramatically as municipalities add the Tesla Robotaxi Network to its infrastructure.

Back of the napkin math: A Tesla taxi could earn $11/hour for 9 hrs a day, 360 days / year. That’s $35k/year. Subtract taxes, depreciation, maintenance, Tesla’s platform fee, and electricity, and there’s around $15k/year in profit. 

Tesla FSD *has the potential* to generate $15k/yr of profit in the future, and is today selling for $8k. 

A NYC taxi medallion sells for $250k-500k. A Tesla with FSD could be similar. 

My Model Y has lifetime FSD. However, it’s high mileage, off warranty, has old AI hardware, tires/taxes/insurance are expensive, the battery is at 80% capacity, and I don’t have home charging. If I got a new Tesla, I could have transferred the FSD token to that one and also received a $7500 federal tax credit.

The software will appreciate in value as the Tesla Network scales its operations, but if Tesla delivers on the Tesla Network at scale, it would be more profitable to put $40k into TSLA the company than a Tesla the vehicle with FSD. 

I predict that the value of FSD will go from $8k today to $80k+ by 2032. And in 2029 TSLA will be valued at $4 Trillion. Holding the stock is free, but holding the vehicle is not. I’m likely selling my Model Y soon. It’s nice but I don’t really need it. I will invest more into Tesla AI.

It’s nice to own a piece of the AI pie.

I’ve learned that one way to get rich is to know everything there is to know about an outstanding company, so that the obvious thing to do is invest a lot and hold for a long time. Ideally forever. But getting rich can be like getting to the starting line. When all the problems are solved, all that’s left is questions, curiosity, and feelings.  

Thanks for reading. Here are some charts I spent hours making.

Thank you

Patrick O'Loughlin

for reading a draft of this at the open feedback gym he hosts for writers.

Drop a comment to have a conversation about Tesla. 

Offerings

Join CrowdHealth, a peer-to-peer healthcare funding network that replaces health insurance and is a superior service. I’ve been a member for three years. It aligns incentives for Crowd Health to make more money when we are healthy. Try it for $99/month.

My first book, All Outcomes Are Acceptable, is for sale on Amazon. It’s about AI drones, Amish people, and getting rich. The reviews are in, readers are loving it.

The Pathless Path Community. This is the best online community I’ve ever been a part of. I recommend joining, especially if you’re an entrepreneur, creator, or curious human not on a default path. It’s a one-time variable payment and you’re in for life.

My publication is free to read but I hope you become a paid subscriber. Financially supporting creators helps create great works of art. With your support, I will become one of the greatest writers in the world.

Click here to support my work.

Watching and Listening

“Optimus is going to be the biggest source of data. It will generate useful data from reality.” 

Music

Thanks for reading, and have a great rest of your day.

Postscript

If Tesla hits 4 trillion dollar valuation, Elon, who owns 25%, would have a trillion dollars of Tesla stock. What will he do with that kind of buying power?

Elon is one of a kind. He has shared many insights publicly and I’ve enjoyed them very much so I decided to begin assembling them into a book. It’s Elon’s Why, How, and What, as well as personal stories and visuals. Let me know if you’d like an advanced copy.

A book I have on the backburner. ETA 2025

I invested in the Model 3/Y program. It changed my life. Somewhat unsurprisingly, the same company is producing more incredible products.

In regard to Optimus, Elon has said something like, “We’re not just starting a new chapter, we’re writing a new book. The Optimus business line will be worth more than everything else at Tesla combined.”

Optimus is basically an embodied supercomputer that will have human-level and superhuman-level capabilities. It’s as significant as the invention of the tractor, airplane, electricity, smartphones, combustion engines, nuclear power.

This will be obvious in hindsight.

Even if a person doesn’t own the rights to the profits of advanced technologies, they can still benefit from increased access and price reductions of goods and services. But my preferred future is one where many people, especially those who work hard to earn it, own a percentage of the profits of advanced technologies.