Rivian Assistant Makes Siri Look Overqualified

[RiverAtlanta]

I think it is a great start and better than Alexa. If Rivian sat on this version and declared it done, it’d be a disappointment, but I am certain they will iterate and improve it like everything else. Glad to to see Rivian innovating like this.

Sponsored

 

[Feetdry]

I found out yesterday that Rivian Assistant doesn't know Ionna. It keeps trying to direct me to EAs, which I don't use these days....

 

[wdwyerR1T]

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.”

I’m more like the uncle, less the margarita, so I’m mildly amused by the experience so far. It IS my first experience conversing (loosely) with someone other than my wife as of late.

 

[Nixapatfan]

I understand that I'm a dinosaur, and that I, unlike others, just don't care about this kind of stuff. But, I'll be dead within 20 years. Carry on.

You're not alone, I'm in the tech industry and still have no desire to have a conversation with my car.

 

[Swiszo]

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.”

I got to "converse" with Rivian assistant yesterday. From the answers I received from her, yes there needs to be a learning period before she can know everything about you. It was sort of like a first date ? After asking what she could control and do and what she could not control and do,I get the feeling that she needs to learn as she goes. Hence the "history" tab on the app.
What I do like right off the bat is that I don not have to take my eyes off the road to turn on or turn off a function in the truck. I'm hoping that it will become better at performing these functions in replacement of me pressings screens tabs that are at many times unresponsive to tapping.
I'm also happy that it appears in this update that the radio stations are now correctly identified and I'm also hoping that my frustration with the favorite seat position not always working and the random changing of the radio stations will perhaps work it way out?
My final thoughts yesterday were that, in my lifetime cars have greatly evolved since the last oil embargo with cool tech and energy savings for the cost of transportation !

By the way, I did ask her how to change her voice and she did say it was possible in the app, but every time I tried I found myself at a dead end !
Cheers !

 

[Captain JB]

It lasted 10 minutes for me…where does the music come from if you cannot use Alexa at the same time? I know I need to read up on it but…….

 

[Zathras]

Another reason to wait for the gen 3 hardware. I'm not letting some agent living at Google or anywhere else, listen to me. Nope. Not a chance. I'll wait for the local language model to give me what I need. I can find my own songs on Tidal in the meantime.

 

[jrmbadger]

I think for me there is a learning curve to figure out how to get the most out of it.

For example, I asked her to turn on the "butt coolers" and she turned on the heated seats! Lol. I had to ask to "turn on the ventilated seats." which worked.

I have been surprised with its capabilities as much as frustrated by some missing capabilities.

For example, I asked it to navigate to Menards (a home improvement chain in the midwest) in city A. The result that came up was city B. I clarified the city I wanted and she came back and told me that "The Menards in City B is commonly referred to as the City A Menards" - and thinking about it more, she was right as City B is a suburb of City A. That was impressive to me.

After that victory for the RA, I asked her to add a stop at a Potbelly's restaurant and it choose one that had nothing to do with my current route. I tried fixing it using RA, but it was a mess, so I started the navigation over. I find the biggest problems are the error handling - e.g., when it gets it wrong, it seems to kinda give up rather than work with you to find a solution.

My worst interaction with it was when I was trying to send a text to my wife. I started talking before it asked me what text I wanted to send (as I wasn't used to the interface) and so the message was kinda messed up. Anyway, it asked me if I wanted to send it, and I said "CANCEL" or something like that and it SENT IT ANYWAY! My wife and I got a chuckle out of that one!

 

[DuoRivian]

Another reason to wait for the gen 3 hardware. I'm not letting some agent living at Google or anywhere else, listen to me. Nope. Not a chance. I'll wait for the local language model to give me what I need. I can find my own songs on Tidal in the meantime.

“Gen 3” hardware aka RAP1 is for autonomy not AI assistant or infotainment

 

[Zathras]

“Gen 3” hardware aka RAP1 is for autonomy not AI assistant or infotainment

I thought RAP 1 was one chip on a motherboard that does it all? And that is Gen 3. Not just the RAP chip.

 

[edwjmcgrath]

I haven't tried Rivian AI but I will say that the experience being described here is not unlike what I see all the time with both Gemini and ChatGPT. I they constantly lose context. I'll ask them to do something and they will do half of it. If I tell them that they didn't do half of what I asked them to do, they treat that statement as informative rather than an instruction to do the other half. If I explicitly tell them to do the other half, they will forget to do the first half. Claude is not perfect but lightyears ahead of the rest. I've taken to getting snarky and thanking Gemini for reminding me why I use Claude for anything important.

 

[SDGriffy]

I’ve found it cannot find and play any of my Spotify playlists. I’ll give it a very specific name and context “play MY Spotify playlist called Sean’s Rush Hour Playlist” and it will find some random playlist that doesn’t even have a similar name.

For navigation test, I asked it to give me directions to the nearest grocery store, it started listing grocery stores and then asked if I would like directions to one in Pasadena, over 100 miles from me. I got angry at it and said “why would I want to go to the farthest option, give me the goddamn closet grocery store to my current location” and it then complied and gave me one a mile away so apparently if you get angry with it, maybe it tries harder?

For controlling climate, ride height and modes, it seems to work great so far. I think they’ll iron out a lot of the kinks with future updates. I’d also like to see 0365 integration so I can receive and reply to work emails, teams messages, etc.

 

[Matchbox]

Bit of a luddite here as I don't normally use voice activated assistant. But for what I wanted to use it for it was very limited - Only knows station numbers and not call letters for the radio (yes - I listen to the radio), When I asked Rivian to fold in my side mirrors it told me to do it myself and how I can do it. When I asked it to vent my windows, it said it could not do that.

So, meh. Maybe it will get better but I have already moved on.

 

[Johnster]

it was documented by Rivian:

It's always a bad idea to expect humans to RTFM!

Sponsored