Rivian Assistant Makes Siri Look Overqualified

[tate16t]

This is an interesting Rivian Assistant issue. See the response from Wassym.

https://www.reddit.com/r/Rivian/s/QSvcb8HKAk

Sponsored

 

[LoneStar]

It sucks big-time. Nothing works right. I get a "blank stare" (wavvy bar goes away without action) like 9 out of 10 times. I tried rephrasing some simple tasks a few ways and it never worked. Not even close. I mean who approved releasing this POS

 

[timgradywy]

I have been surprised in a good way by RA so far. It's done everything I asked and even what my wife asked. It's dramatically better than Alexa ever was for me. I love being able to change ride height and modes by voice instead of having to do a bunch of touching around. The texting features are also nicely implemented. For a first pass, I think they did a great job.

 

[ENVErider]

Some of you must be a barrel of fun at other peoples' birthday parties...

It just came out and will get updates and improvements. This was a game changer for me on my R1S. Use it quite a bit. Never used Alexa, Siri, or Google Music so I am not looking at through that lens.

Texting works great. Nav works great. I pull out and say "Hey Rivian shut my garage door" in one sentence and it immediately shuts. Say "Hey Rivian take me to the closest Starbucks and then work" in one sentence - it confirms the nearest Starbucks and then satnavs the trip. "hey Rivian, turn off passenger side vents" and it flawlessly does. I am not seeing the delay. Again, it is not perfect, but what is? Would I like it to control the wipers and such - yes - but it does not do that. No reason for me to yell at the clouds.

This is obviously a "tough crowd" making me glad that I'm not a Rivian engineer reading this thread. I agree with all you said and even tried the wiper control too, but I wasn't surprised or disappointed, given it's a physical stalk-controlled item.
I'm thrilled with my new RA features; my son and I are having fun trying new things as we learn all RAs capabilities.

 

[Virtio]

It's sooooo much better than Alexa and much more fun too. I've been playing games, getting tips on dining and local events, movie times, and much more. She's quite the chatterbox and is quite complimentary too.

 

[klaopune]

I genuinely appreciate Rivian trying to bring AI into the vehicle experience. That’s ambitious. That’s forward-looking. That’s the kind of thing that makes you feel like you’re living in the future.

Unfortunately, after using Rivian Assistant for several days, I can confidently say the future apparently runs on a 2017 Bluetooth speaker and mild confusion.

I’m honestly baffled by how far behind this feels compared to literally every modern LLM on the market. Claude, ChatGPT, Gemini — all of them can hold context, remember what you just said, and generally behave like they’re awake.

Rivian Assistant, meanwhile, has the memory retention of a golden retriever in a laser tag arena.

Example:
Me: “Find that song I like.”
Rivian Assistant: finds it
Me: “Great, play it.”
Rivian Assistant: “Absolutely. Here’s a completely different song from three playlists ago that nobody asked for.”

It’s like talking to a very enthusiastic intern who got hit in the head with a Roomba.

And this isn’t some obscure edge case. This is basically the entire experience. Every interaction feels like the assistant is waking up from a nap, hearing one random noun, and making a panic decision. On that note, I'm going to take a break from this buggy tech and go enjoy a smooth win with a mostbet promo code bangladesh instead.

At this point I’ve gone back to Alexa, which feels like admitting defeat in a futuristic society. That’s like buying a spaceship and then commuting via horse because the autopilot keeps driving into lakes.

What confuses me most is: why reinvent the wheel here?

Why not integrate Claude or another mature LLM provider? Because right now this feels less “AI-native EV experience” and more “two Raspberry Pis zip-tied together behind the dashboard running Ask Jeeves.”

Surely the token costs can’t be THAT bad. I refuse to believe we’re rationing GPU cycles like wartime butter.

And the frustrating part is that the vision is actually great. An intelligent in-car assistant should be a killer feature for Rivian. This SHOULD be the perfect environment for contextual AI.

Instead, my truck currently has the conversational abilities of a haunted GPS unit.

I want this to succeed. I really do. But today, Rivian Assistant feels less like “the future of driving” and more like your uncle discovering ChatGPT after two margaritas and saying “watch this.”

The current mess with RA comes down to how the system is built. Right now, it relies on a hybrid model where an orchestration layer has to constantly decide which commands to process locally in the car (like rolling down windows or changing drive modes) and which ones to send up to Rivian's private cloud (like searching for music or general knowledge).

When you say smth like "play that song I like" the cloud LLM might understand your intent, but when it passes that token back to the local vehicle media API, the context gets stripped or misaligned. The car's local system treats it as a brandnew, isolated command instead of part of a continuous conversation, which is why it panics and defaults to a completely different playlist.

As for how fast this will get fixed, there are two different timelines here:

For existing R1 vehicles: Since the feature just dropped in the 2026 OTA update, the immediate cloud to local context handoff bugs will likely see hotfixes within the next few software cycles. However, because Gen 1 and Gen 2 R1s rely heavily on cloud processing for the complex LLM logic, it will always be bottlenecked by cellular latency and cloud to car API translation.

The real fix is the assistant won't feel truly seamless until it runs entirely locally. Rivian is debuting a much more powerful 200 TOPS onboard Edge AI compute system on the new R2 platform, which is designed to run a smaller LLM natively inside the vehicle hardware.

Until that heavy local compute architecture matures and tricks down or gets heavily optimized for the current R1 fleet, we are basically beta testing an orchestration layer that is still learning how to talk to the car's media player

 

[Dark-Fx]

In fairness to the OP, I'm thinking the issue is less about Rivian's shortcomings and more about why they chose to release a homemade, yet subpar feature, when the baseline is already established in the field. This is akin to releasing a new CD player that also reads .mp3s into a world that's already full of established multimedia players. Perhaps there's proprietary barriers to folding mature AI instead of homegrown into the Rivian? Just spitballing.

I'm not interested in fighting about this. Just pointing out my interpretation.

A big part of the OP's initial complaint stem from not enabling one of the more important features required for what they wanted to use it for. They also didn't seem to know it's built on Gemini because they suggested it be used.

I wouldn't be surprised if a lot of the complaints stem from not turning on the memory option. While it doesn't seem to remember everything from drive to drive, it does pretty good about follow-up questions to something that was already said or asked on that same drive.

My main complaint is the single calendar thing. I'm guessing that might be more of a Google problem that Rivian will need to work around though. If I actually intended on using it to the full extent of usefulness,I'd have to do some third party integration that keeps the main calendar synced with all my extra calendars. I use the calendars by basically having one set up per topic so it's easier to manage between people that share these calendars.

 

[Noplacelikeloam]

So, it’s been a few days now with the memory option turned on. Well, there is some notable improvement, especially that I can now begin to train it on some basic preferences, but the latency is not useful, many requests still with no reply whatsoever, it doesn’t work well if the passenger or anyone else in the car is talking to it, you have to speak very clearly for it to understand what you’re saying, and at this point, I’ve just gone back to asking Siri to do most tasks. I have to tip my hat to the Rivian team for getting this to market so quickly, but in hindsight, wish they had spent a few more months making this the product it should’ve been.

 

[electruck]

So, it’s been a few days now with the memory option turned on. Well, there is some notable improvement, especially that I can now begin to train it on some basic preferences, but the latency is not useful, many requests still with no reply whatsoever, it doesn’t work well if the passenger or anyone else in the car is talking to it, you have to speak very clearly for it to understand what you’re saying, and at this point, I’ve just gone back to asking Siri to do most tasks. I have to tip my hat to the Rivian team for getting this to market so quickly, but in hindsight, wish they had spent a few more months making this the product it should’ve been.

poor customer perception is the inherent danger of an MVP launch and learn approach vs taking a lot longer trying to perfect what you think the customer wants before releasing it. welcome to the new software defined world where the customer is key to product development.

 

[Noplacelikeloam]

poor customer perception is the inherent danger of an MVP launch and learn approach vs taking a lot longer trying to perfect what you think the customer wants before releasing it. welcome to the new software defined world where the customer is key to product development.

And so many examples of how this is gone, terribly wrong, Google, glass, Apple vision, the launch and learn approach, is always curious to me. Launch small and scale gradually would’ve been a better approach.

 

[HaveBlue]

I cannot get it to recognize or access any of my contacts. It is driving me insane. I see the contacts, they are loaded, but the assistant does not. (Also as a aside, this voice is sooo annoying.)

On an android phone you have to agree to a pop up (toast) on your phone. Miss that and it won't work.

 

[TexasBob]

On an android phone you have to agree to a pop up (toast) on your phone. Miss that and it won't work.

Thanks. All my contacts sync to the truck (and I can see it on the screen), but the assistant cannot see any of them.

 

[Supratachophobia]

I genuinely appreciate Rivian trying to bring AI into the vehicle experience. That’s ambitious. That’s forward-looking. That’s the kind of thing that makes you feel like you’re living in the future.

Unfortunately, after using Rivian Assistant for several days, I can confidently say the future apparently runs on a 2017 Bluetooth speaker and mild confusion.

I’m honestly baffled by how far behind this feels compared to literally every modern LLM on the market. Claude, ChatGPT, Gemini — all of them can hold context, remember what you just said, and generally behave like they’re awake.

Rivian Assistant, meanwhile, has the memory retention of a golden retriever in a laser tag arena.

Example:
Me: “Find that song I like.”
Rivian Assistant: finds it
Me: “Great, play it.”
Rivian Assistant: “Absolutely. Here’s a completely different song from three playlists ago that nobody asked for.”

It’s like talking to a very enthusiastic intern who got hit in the head with a Roomba.

And this isn’t some obscure edge case. This is basically the entire experience. Every interaction feels like the assistant is waking up from a nap, hearing one random noun, and making a panic decision.

At this point I’ve gone back to Alexa, which feels like admitting defeat in a futuristic society. That’s like buying a spaceship and then commuting via horse because the autopilot keeps driving into lakes.

What confuses me most is: why reinvent the wheel here?

Why not integrate Claude or another mature LLM provider? Because right now this feels less “AI-native EV experience” and more “two Raspberry Pis zip-tied together behind the dashboard running Ask Jeeves.”

Surely the token costs can’t be THAT bad. I refuse to believe we’re rationing GPU cycles like wartime butter.

And the frustrating part is that the vision is actually great. An intelligent in-car assistant should be a killer feature for Rivian. This SHOULD be the perfect environment for contextual AI.

Instead, my truck currently has the conversational abilities of a haunted GPS unit.

I want this to succeed. I really do. But today, Rivian Assistant feels less like “the future of driving” and more like your uncle discovering ChatGPT after two margaritas and saying “watch this.”

The analogies in this post are the real heroes.

 

[jemkewl]

Turned on Rivian Assistant. It is off now. Alexa works better for navigation and weather which is all I need. Rivian Assistant begins responding and then mid-sentence and mid-word just stops. That is not a one time thing, it happens every time. This is not even an alpha release.
Might try again if Amazon Music becomes available and Rivian at least gets to a beta release.

 

[electruck]

Turned on Rivian Assistant. It is off now. Alexa works better for navigation and weather which is all I need. Rivian Assistant begins responding and then mid-sentence and mid-word just stops. That is not a one time thing, it happens every time. This is not even an alpha release.
Might try again if Amazon Music becomes available and Rivian at least gets to a beta release.

That, specifically, has not been my experience at all. Are you in an area with poor reception or cellular data (the latter should not to be confused with the number of signal bars and can be caused by network congestion, independent of signal strength)? A lot of the processing is happening off-vehicle so wondering if that might be causing your problems. Doesn't make it any more usable but it's a different problem than being buggy.

Sponsored