Rivian Autonomy & AI Day! Lidar, Hands-Free Driving, Rivian Assistant & more announced

[BigSkies]

Rebound in the stock after initial profit-taking; could be there's support for the strategy...

Morgan Stanley downgraded the stock today. At least if I trust the Yahoo Finance AI summary of such things.

This has more to do with the price movement than anything autonomy related.

Sponsored

 

[jrmbadger]

Maybe I'm an outlier, but all I really need is lane steering and auto lane change (e.g., enhanced autopilot) to help with long drives. I like making the decisions by myself in the car. That's why I'm getting a comma for my gen 1 R1S - it has basically all I want and is a one time fee.

I've tried FSD on my Model Y when we get periodic free trials. I've stopped accepting the free trials as the whole thing gives me anxiety as I sit at an intersection trying to guess whether a computer that can react thousands of times faster than I can is going to pull out in front of a car at the last minute and I have to jump in and save my family. Maybe I would feel differently if Tesla would accept responsibility for the crash, but the fact that Tesla wont put its money where its proverbial mouth is doesn't give me a lot of confidence in its product.

 

[Dark-Fx]

What they actually demonstrated today was FSD 10 levels of driving. That's not good

What FSD version was Tesla at before you were like "This is it, they never have to improve beyond this for unsupervised full self driving"?

 

[Zorg]

Developing their own chip seems quite the endeavor though time will tell. Otherwise, excited about where they're going.

 

[Hereforthesnacks]

RJ just killed off R1. R2 is gen 3 including LiDAR and Rivian developed AI chip.

This is exactly the reason holding onto EVs for more than 3~5 yrs is just like computers - outdated.

Just as I've said before these vehicles are laptop with wheels and cannot place ICE paradigm onto EVs.

You get the same rapid upgrades in tech for ICE cars. Any new, high tech car, ICE or EV, is obsolete in 5 years (or at least will stop getting meaningful updates). This is not an EV specific problem.

 

[sunydrm]

What FSD version was Tesla at before you were like "This is it, they never have to improve beyond this for unsupervised full self driving"?

None. Because FSD 14 needs to improve to have unsupervised

It got past the gimmick stage around FSD 12.5.6 and then 13.2.9 was ultra smooth.

Rivian might have more features than tesla had in a similar amount of time(like parking spot to parking spot) .

HW3 was around 120 TOPS and 8gb of memory. Rivian will have 1600 and much more memory

 

[Dark-Fx]

None. Because FSD 14 needs to improve to have unsupervised

Then your entire point is meaningless.

 

[Hereforthesnacks]

No bounce from this yet:

Well, the news isn’t exactly mind blowing. Nothing unveiled today would push the stock up. It’s a catch up game with Rivian. But they have to do it with 0 profit (for now?).

 

[Hereforthesnacks]

Very solid tech stack. Here is a summary from Gemini of its competitive positioning fwiw.

Here is the summary of the Rivian Autonomy Stack and where it stands competitively.

1. The Rivian Autonomy Stack (The "Peregrine" Architecture)
A. The Hardware: "Thor-Class" Custom Silicon Rivian isn't buying chips; they built a monster.

• The Chip: RAP1 (Rivian Autonomy Processor 1).
• Specs: 1,600 TOPS (Sparse INT8) per Gen 3 computer (Dual Chip). This is massive.
• The "Killer Feature": The Neural Engine has a dedicated hardware block for "Unstructured Point Cloud Processing".
  • Significance: Unlike Tesla (which forces everything into 2D video), Rivian’s chip eats raw Lidar/Radar data natively. This is hardware-level validation of the "Multimodal" thesis.
• Process: TSMC 5nm Automotive.

B. The Software: "Vision-Language-Action" (VLA) They are explicitly moving to the reasoning architecture

• Architecture: Sensors → Transformer Encoders → World Model → Trajectory.
• Training Tech: They explicitly mention "Tokenized Data" and "GRPO Learning"(Group Relative Policy Optimization).
  • Significance: GRPO is a cutting-edge Reinforcement Learning technique used to train LLMs to "reason." Rivian is applying this to driving, confirming they are treating the car like an agent, not a reflex loop.
• Validation: "Apprentice Mode". This is their "Shadow Mode"—the agent runs in the background, predicting moves and learning from the human driver without taking control.

2. Competitive Analysis

Feature	Rivian (Gen 3)	Tesla (HW4 / AI4)	China (Zeekr/XPeng)	Verdict
Compute	1,600 TOPS	~300–500 TOPS	2,000+ TOPS(Thor/Turing)	Rivian ≈ China > Tesla. Rivian matches the high-end "Thor" standard 18 months early.
Sensors	Lidar + Radar + Cam	Camera Only	Lidar + Radar + Cam	Rivian Wins on Robustness. The slides explicitly show Lidar detecting obstacles in fog where cameras fail.
Architecture	World Model / VLA	End-to-End NN	End-to-End NN	Parity. Everyone is converging on World Models, but Rivian has the compute to run it now.
Scalability	Chiplet (RivLink)	Single SoC	Single/Dual SoC	Rivian Wins. The 128 Gbps interconnect allows them to stack chips easily.

Export to Sheets
3. The "Secret Sauce" Approach
Rivian’s presentation reveals three strategic pivots that separate them from the pack:

1. Hardware-Native Fusion: They didn't just add Lidar; they built a chip that speaks Lidar. This eliminates the processing penalty that usually makes sensor fusion slow.
2. "Agentic" Cabin: They are moving to Edge AI Agents. The slides show a "Small Language Model" (SLM) running on the chip (not in the cloud) to handle voice and vehicle commands. This means your car is smart even when offline.
3. The "Data Recorder" Edge: Their "Autonomy Data Recorder" specifically identifies edge cases (animals, tunnels, weird merges) on the car and uploads only that interesting data. This solves the "Data Deluge" problem—they don't upload boring highway miles, just the confusing stuff to train the World Model.

Summary: The "Prophet" Validation
Rivian has effectively positioned itself as the "Western Waymo"—a consumer brand that owns the full "Chip-to-Wheel" stack with the sensor robustness of a Robotaxi and the business model of a luxury automaker.

I’d like to see the prompt that lead to this slop lol

 

[sunydrm]

Then your entire point is meaningless.

They have been working on this several years. They shortcutted tesla in using end to end neural networks and have a computer that is 5-10x more powerful than what Tesla is using with HW3. They also have more sensors.

My point is these sensors or end to end AI do not bring autonomy faster

You can look at the progress of chinese systems. They get to FSD-like capabilities very quickly. The safety and driving quality appears to stagnate because the initial driving abilities come very quickly

 

[tkauftheil]

My 2 cent take away from their announcement and reading thru this thread:

FSD is neat, but I’m not all that interested. The current driver+ I have (g2 R1S) is plenty sufficient. Autonomy so I can do other things in the car also doesn’t interest me. Unlike many others, any of my commutes is me time to relax and listen to music or audio books. Certainly not work. Might be helpful to be able to eat and drive, but I don’t want the mess in my car. Plus I’ve learned to slow myself down so not to be rushed anymore.

Hands free texting and reading of texts. I have that now and it works great. Bt my phone, ask Siri to read or compose. Not sure what everyone is so hung up on. Do you not bt your phone to your car. I can also answer calls and make calls. I rarely keep my phone in my pocket so I never encounter the unlocking issue. I have a phone mount that works perfect.

Obsolescence. Not a huge concern for me personally as I lease. And I leased on purpose so that I can “upgrade” every couple of years. The Rivian is more technology than an automobile. Leasing made sense to me.

R2 early and late versions. I originally got into my R1 as a stepping stone into an EV and as a precursor to the R2. However I fell in love with the size and space of the R1. Going smaller might be a hard sell for me. Although I do plan on moving to Europe in the next few years, and if the R2 is available over there, I’d get one for sure.

oh, back to full self drive, I don’t get the lure of it. If you don’t want to actually drive, get into a ride share life style.

End of my 2cent tour of my brain on this.

 

[VandalSibs]

HA HA HA That's a horrible take and/or justification for your decision. These are on average > $90K depreciating assets, if you know in 6 months that your car will be outdated, why would you buy one today, knowing the price is staying roughly the same and the perceived value is vastly improved.

I bought my Gen 1 for what it is, but believed in a company making huge promises, I don't want or need the features of Gen2 or 3 - I just want a good ADAS system and UX/UI that isn't full of bugs.

If the vehicle does what I need it to do when I buy it, it's worth the purchase to me. Not everyone gets a product for the potential features, if it already does what they need it to do. Anything else is just the cherry on top (I don't like cherries on ice cream all that much, but it fits).

When I was shopping for a truck, I look at the following use requirements and needs:

1. Be an EV.
2. Fit in my wife and I's budget.
3. Fit in our garage.
4. Be able to tow our trailer further than our previous vehicle (Model Y)
5. Have enough space to do the yard projects my wife and I take on.

Guess what? My Gen 1, 2023 (bought in 2024) R1T fulfilled all of those requirements! And it did it without self-driving, the ability to send text messages, hover, etc. It does all the things that I need it to do, so therefore we got it.

Like, I get that some people do the analysis on value depreciation and incorporate it into the purchasing plan. But for many, many others, they just want a vehicle that can do what they want it to do.

 

[Dark-Fx]

Website has some more info that I don't recall Rivian touching on exactly.

And an answer to the "lifetime" upfront cost:

5. Lifetime upgrade product features remain available during the lifetime of feature support for the hardware on the vehicle at delivery. Lifetime upgrades remain with the vehicle upon ownership transfer.

 

[blinkidy]

So with everything we know do we think we can expect automatic lane change before point to point? Was kind of expecting auto park and auto lane change to be announced with universal hands free

 

[Jeff B.]

Holding RIVN stock is like riding a roller coaster. Lots of high expectations, ups / downs and queezy moments. I have managed to breakeven buying and selling the up / downs and I don't mind continuing to take small profits on these daily moves, but R2, in whatever form, can't get to market soon enough.

Sponsored