Ben Agro

Trained FSD v14 @Tesla_AI, 🔜 Robotaxi . ex @Waabi_ai, @UofTCompSci and @VectorInst. Rock climber 🧗

[Ben Agro @BenAgro4]

Huge congrats to the team! Exciting milestone, but only the beginning!

[Phil Duan @pduan]

Hi there!

[legal YIMBY alien @YIMBYPoland]

Translation: if you’re a advanced researcher at OpenAI on the O1 visa (extraordinary ability) and you apply for a green card, they’ll make you go back abroad and wait years for an appointment. Of course, by then you probably no longer work at OpenAI.

[Tesla Robotaxi @robotaxi]

Robotaxi now rolling out in Dallas & Houston 🤠

[Elon Musk @elonmusk]

Just started Tesla Robotaxi drives in Austin with no safety monitor in the car.

Congrats to the @Tesla_AI team!

If you’re interested in solving real-world AI, which is likely to lead to AGI imo, join Tesla AI. Solving real-world AI for Optimus will be 100X harder than cars.

[Ashok Elluswamy @aelluswamy]

Robotaxi rides without any safety monitors are now publicly available in Austin.

Starting with a few unsupervised vehicles mixed in with the broader robotaxi fleet with safety monitors, and the ratio will increase over time.

[TSLA99T @Tsla99T]

I am in a robotaxi without safety monitor

[roon @tszzl]

@hopes_revenge only car that drives itself. which unfortunately makes it a completely different product category than all the other cars

[Oliver Cameron @olivercameron]

5 years later and I think @Tesla has legit cracked FSD.

Tried the latest version today and it was genuinely flawless, and I critique every detail of anything self-driving.

Not only flawless, but it drove smoother and safer than any human I know.

Very impressed.

[Ashok Elluswamy @aelluswamy]

@elonmusk It’s an amazing experience!

[Jim Fan @DrJimFan]

I was very late to own a Tesla but among the earliest to try out FSD v14. It's perhaps the first time I experience an AI that passes the Physical Turing Test: after a long day at work, you press a button, lay back, and couldn't tell if a neural net or a human drove you home.

Despite knowing exactly how robot learning works, I still find it magical watching the steering wheel turn by itself.

First it feels surreal, next it becomes routine. Then, like the smartphone, taking it away actively hurts. This is how humanity gets rewired and glued to god-like technologies.

[Phil Duan @pduan]

Along for the ride in unsupervised FSD testing

[Sawyer Merritt @SawyerMerritt]

NEWS: A second Tesla Model Y Robotaxi running FSD Unsupervised has just been spotted driving itself on public roads in Austin, Texas, with no one in the front seats.

This is a different car from the one spotted earlier. They have different license plates.

h/t @Mandablorian

[420 Bounty Hunter @Mandablorian]

Just saw a Tesla with no one in the car @SawyerMerritt @wholemars @robotaxi @Tesla

[Pete Balls to the wall FSD 🤖🚕 @kylaschwaberow]

This might be the wildest clip from a @wholemars video ever.

I can’t stop laughing 😂😂🤣

[Whole Mars Catalog @wholemars]

Tesla Self-Driving 14.2 was sent from the heavens to the Earth, from the gods.

It must be so, because only gods have the power to create life. To bestow an inanimate body with the ability to think, to react, and to move on its own.

The weights were sent down from the heavens to save our lives, to save our pets, to save the wildlife, and to preserve the Earth.

Much credit to the Tesla AI team, but this software only flowed through them. The impact is so huge, the results are so shocking, that this must be divine.

[Andrej Karpathy @karpathy]

I took delivery of a beautiful new shiny HW4 Tesla Model X today, so I immediately took it out for an FSD test drive, a bit like I used to do almost daily for 5 years. Basically... I'm amazed - it drives really, really well, smooth, confident, noticeably better than what I'm used to on HW3 (my previous car) and eons ahead of the version I remember driving up highway 280 on my first day at Tesla ~9 years ago, where I had to intervene every time the road mildly curved or sloped. (note this is v13, my car hasn't been offered the latest v14 yet)

On the highway, I felt like a passenger in some super high tech Maglev train pod - the car is locked in the center of the lane while I'm looking out from Model X's higher vantage point and its panoramic front window, listening to the (incredible) sound system, or chatting with Grok. On city streets, the car casually handled a number of tricky scenarios that I remember losing sleep over just a few years ago. It negotiated incoming cars in tight lanes, it gracefully went around construction and temporarily in-lane stationary cars, it correctly timed tricky left turns with incoming traffic from both sides, it gracefully gave way to the car that went out of order in the 4-way stop sign, it found a way to squeeze into a bumper to bumper traffic to make its turn, it overtook the bus that was loading passengers but still stopped for the stop sign that was blocked by the bus, and at the end of the route it circled around a parking lot, found a spot and... parked. Basically a flawless drive.

For context, I'm used to going out for a brief test drive around the neighborhood to return with 20 clips of things that could be improved. It's new for me to do just that and exactly like I used to, but come back with nothing. Perfect drive, no notes. I expect there's still more work for the team in the long march of 9s, but it's just so cool to see that we're beyond finding issues on any individual ~1 hour drive around the neighborhood, you actually have to go to the fleet and mine them. Back then, I processed the incredible promise of vehicle autonomy at scale (in the fully scaleable, vision only, end-to-end Tesla way) only intellectually, but now it is possible to feel it intuitively too if you just go out for a drive. Wait, of course surround video stream at 60Hz processed by a fully dedicated "driving brain" neural net will work, and it will be so much better and safer than a human driver. Did anyone else think otherwise?

I also watched @aelluswamy 's new ICCV25 talk last week (https://t.co/RdaM23kvez) that hints at some of the recent under the hood technical components driving this progress. Sensor streams (videos, maps, kinematics, audio, ...) over long contexts (e.g. ~30 seconds) go into a big neural net, steering/acceleration comes out, optionally with visualization auxiliary data. This is the dream of the complete Software 1.0 -> Software 2.0 re-write that scales fully with data streaming from millions of cars in the fleet and the compute capacity of your chip, not some engineer's clever new DoubleParkedCarHandler C++ abstraction with undefined test-time characteristics of memory and runtime. There's a lot more hints in the video on where things are going with the emerging "robotics+AI at scale stack". World reconstructors, world simulators "dreaming" dynamics, RL, all of these components general, foundational, neural net based, how the car is really just one kind of robot... are people getting this yet?

Huge congrats to the team - you're building magic objects of the future, you rock! And I love my car <3.

[Igor Babuschkin @ibab]

A common mistake that AI companies make nowadays is to not give their engineers enough time and mental calm to do their best work. Constant deadlines, pressure and distractions from daily AI news are poison for writing good code and systems that scale well. That’s why most AI APIs and products have reliability issues.

A good company culture that mixes excellence with focus and enough rest leads to faster and better results. The best example of how to do it well is the early Google culture from 1998 which resulted in one of the largest scale and most reliable services on the web in just a few short years. Founders should copy some of the strategies that Larry and Sergey used. They are still underrated IMO despite their huge reputation.

[Ashok Elluswamy @aelluswamy]

https://t.co/Jr4i7tHUCr

[Matthew Wilson @matwilso]

Andrej's outputs feel like whatever the opposite of slop is

Great ep. I especially appreciated the counterbalance to all the "fundraising" hype and uninformed, shallow takes on the timeline. Made me feel less like I'm on crazy pills.

The points on self-driving resonated. It's not solved yet. Waymo driverless deployments are great, but still small and not economical. The driver has just been moved to somewhere we can't see them lol. We have marched many 9s from initial demos, but have more to go. Tesla seems on track to scale better, and IMO is uniquely well suited to create tons of value.

Worth a (re)watch. High-quality tokens are rare

[Mike P @mikepat711]

V14 - 35 highway miles in mad max mode. It is very aggressive but simultaneously feels way safer than even standard mode V13 highway. All of my former highway complaints are cured. Hyper-aware of traffic pattern changes now, follow distance way better even in mad max. Much more confidence inspiring. Also zero brake stabs on this drive.

I got off the highway and had to merge into super dense traffic. The cut in on mad max was incredible. Honestly better at being assertive and safe than almost all human drivers. No way anyone but a pro is competing with the mixture of assertiveness and safety.

I’m getting so hype rn omg. This feels like the big one. I so far see zero reason why this can’t go unsupervised inside of the Robotaxi fleet at the very least.

Do not bet against Tesla AI fam. You’re gonna get smoked.

[Zack @BLKMDL3]

10 drives in with FSD v14.1 and here are all my thoughts.

First of all- wow. Zero disengagements or interventions thus far. The confidence and overall human like driving is next level. The steering inputs are so smooth, braking inputs are earlier and more linear than before.

It gives off Robotaxi vibes with how it navigates parking lots, plus how it drives and how quickly it backs into a parking spot. Such a nice experience to ride along with it.

Seeing FSD now navigate parking garages is so cool. When you enter it waits for the arm to open after you get your ticket, proceeds in and finds a parking space. When you leave it does the same thing, gets close to the ticket stand and then proceeds. Really fascinating to see in person.

It really sometimes feels like FSD can now read signs in the garage to guide it to the exit. Brought it to the back corner of a garage and it found its way out.

Hurry mode is awesome, quick and more assertive than before, while also being smoother with less unnecessary lane changes. Sloth mode is what you’d expect- exact speed limit and gentle driving. Icons are cool and easy to change. When you start FSD from park, there’s no delay. It starts right away and leaves in a split second. Really good improvement.

I love being able to select the Arrival Options if I am parking curbside or in a charger, it also automatically chooses depending on where you navigate to.

It perfectly parks at Superchargers, tried it at the Tesla Diner and at the Santa Monica supercharger and both times it was excellent. Always centered in the lines and it parks better and quicker than the majority of humans do.

Only thing so far to note is one instance of slight braking when going around a bus blocking a lane, was probably not even 1/2 of a second and dropped 1-2mph but felt it. Super super minor.

Had FSD v14 move over for construction, a loose cone on a dark road, and obey a worker holding a stop sign. All of them were smooth and felt human like or better.

It pulls in and out of my driveway great, doesn’t hesitate at all as well as when at chargers. Everything has a quicker response time. Curbside is cool too as it pulls right up against the curb perfectly in a parking spot.

This release now means now 100% of your driving can now be done on FSD. From your driveway into a parking garage. Elon was right when he said it would feel sentient, and this isn’t even v14.2 yet.

The Tesla AI team COOKED with this update. It’s phenomenal. Huge congrats to them for such an epic release. This is a HUGE update and I can’t wait to drive it more.

It’s now 5am so I’m going to attempt to get some sleep, but tons more driving and videos coming later today as soon as I can. Thanks everyone for following along.