Toggle sidebar Toggle sidebar
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

Grok AI Voice Mode Arrives on Apple CarPlay

Does it call itself Grok or... You know. Its concerning name.

But in all seriousness, ok. I won't use it. The most I use siri for is to pull up stuff on ye maps or call someone. Maybe text. But I'm sure there are people who will find uses for these type of things. And as cars go more automated, sure.
 
decimortis said:
I ask this in all seriousness, but can someone tell me a true, real-world advantage to having a chatbot in my car with me? I watched the little Grok video in the article for a potential answer. My question still stands.
Click to expand...
It’s integrated into the Tesla navigation system. “Set a waypoint to the closest Wawa”. “Hey Grok, what are some restaurant recommendations on my current route”. “Hey grok what are the reviews on this restaurant”.
 
  • Like
Reactions: BigMcGuire
Veinticinco said:
Struggle with the same question.

Maybe random vey isolated needs like if you’re running late for a meeting and need to know if multi-story parking spaces are available without taking a chance on alternatives, or wasting time pulling over to check on your phone.

Once a year at most.
Click to expand...
Well you don’t have to use the app except for once a year.
 
  • Like
Reactions: BigMcGuire
tech_enthusiast_ said:
I'm sorry but this is just so depressing. 5 year olds have been asking questions that parents don't know the answer to since the dawn of time and humans before us didn't need an AI chatbot to engage with their child. You're not encouraging curiosity in your child if you teach them to rely on AI to answer every question they have without doing any critical thinking to find the answer.

If you don't know the answer to their question, a better way to encourage your child's curiosity is to be honest with them and say "I'm not sure, but we can figure it out when we get home." Then engage with them in critical thinking to find the answer without just giving it to them. This also teaches them that you can't always get the answer to your questions instantaneously.
Click to expand...
I disagree. There are simple questions that for most of us we won’t be able to give anything other than a surface answer. “Hey Grok: tell me the history on the flatiron building”.
 
Icehawk said:
Does it call itself Grok or... You know. Its concerning name.

But in all seriousness, ok. I won't use it. The most I use siri for is to pull up stuff on ye maps or call someone. Maybe text. But I'm sure there are people who will find uses for these type of things. And as cars go more automated, sure.
Click to expand...

I don't have Grok in my Tesla because mine is running a little older hardware and does not support it. I have demo'd a Tesla with it and yeah not sure there is a lot I would actually use it for if it was not able to replace Siri and give directions and things. As a secondary AI assistant it seems limited and think we are still at the dawn of what AI will actually do that will change how we interact with technology. At some point it will be our own personal Jarvis like Tony Stark had in the Marvel movies. Seamlessly moving from your house to your phone or to your car or other objects, carrying over all your preferences, knowing your schedule and everything else. No more reminder apps and things like that, but an AI assistant that would be like talking to a real life assistant.
 
tech_enthusiast_ said:
I'm not sure you read what I wrote
Click to expand...
I think your idea is a bit utopian and not realistic sadly. AI is no different than googling it, just a faster and smoother interaction. Few are going to be cracking open the books to find out the history of a building you are passing by, they are going to google it or now they are going to AI it. What it does is allow you to engage in conversation immediately rather then waiting till you get home. So AI will give the factual details (hopefully) and then you can lead into a discussion about the rest. Treat it as an aid rather than a parent replacement.
 
  • Like
Reactions: BigMcGuire
wonderings said:
I am curious, do you drive a Tesla? I have had one through work for about a year now and would not want to go back to CarPlay. The whole system is butter smooth and works really well, CarPlay would look incredibly out of place and childish in the Tesla eco system.
Click to expand...
Not who you're responding to, but I do drive a Tesla and miss CarPlay for certain things. 1)Group messaging on iMessage 2) audio apps like overcast/mlb 3) Waze/Maps (Tesla's are pretty good, but sometimes it makes weird decisions for me and I like to have a second opinion. 4) Apple Music (At least on my HW3 car, the built in Apple Music app is slow and, in my opinion, poorly laid out). Definitely better than connecting via bluetooth, but I find the CarPlay implementation much better.

I think they could make it an app and people could just use as needed. I definitely wouldn't have it up 24/7, but I'd like the option when I needed. It would also mean I could get rid of the phone mount I have, which is ugly and I dislike. But totally understand why someone would think it's not needed - Tesla is miles and miles ahead of any other car infotainment system I've tried and definitely does some things better than CarPlay.
 
  • Love
  • Like
Reactions: BigMcGuire and turbineseaplane
Icehawk said:
Does it call itself Grok or... You know. Its concerning name.
Click to expand...
I don’t know does Siri call itself Siri.
Icehawk said:
But in all seriousness, ok. I won't use it. The most I use siri for is to pull up stuff on ye maps or call someone. Maybe text. But I'm sure there are people who will find uses for these type of things. And as cars go more automated, sure.
Click to expand...
It’s beats asking my wife to use her phone to find a restaurant en route . I can’t wait until I can control my car with grog and ditch the strangely worded voice commands.
 
  • Like
Reactions: BigMcGuire
surferfb said:
Not who you're responding to, but I do drive a Tesla and miss CarPlay for certain things. 1)Group messaging on iMessage 2) audio apps like overcast/mlb 3) Waze/Maps (Tesla's are pretty good, but sometimes it makes weird decisions for me and I like to have a second opinion. 4) Apple Music (At least on my HW3 car, the built in Apple Music app is slow and, in my opinion, poorly laid out). Definitely better than connecting via bluetooth, but I find the CarPlay implementation much better.

I think they could make it an app and people could just use as needed. I definitely wouldn't have it up 24/7, but I'd like the option when I needed. It would also mean I could get rid of the phone mount I have, which is ugly and I dislike. But totally understand why someone would think it's not needed - Tesla is miles and miles ahead of any other car infotainment system I've tried and definitely does some things better than CarPlay.
Click to expand...
Open for all to discuss! I did forget about messaging and definitely not as nice as what is in CarPlay. Slipped my mind because I just don't do a lot of texting and if I do it does pop up and read it to me. Trying to remember if I can reply hands free as well.... shows how much I need it personally!

I like Tesla's maps, I used Waze all the time on my previous car with CarPlay and am happy with what Tesla has to offer. Apple Music runs well for me and also running HW3, much smoother then CarPlay for sure and reminds me of the App for MacOS and with the little navigating in the app I do it works great.

I am not against options, just personally don't see CarPlay as an improvement over all that would have me running the colorful circus that is CarPlay. I am surprised that Tesla has not opened itself up to more apps being created rather than them just curating what they want on the car.
 
  • Like
Reactions: BigMcGuire, I7guy and surferfb
wonderings said:
I think your idea is a bit utopian and not realistic sadly. AI is no different than googling it, just a faster and smoother interaction. Few are going to be cracking open the books to find out the history of a building you are passing by, they are going to google it or now they are going to AI it. What it does is allow you to engage in conversation immediately rather then waiting till you get home. So AI will give the factual details (hopefully) and then you can lead into a discussion about the rest. Treat it as an aid rather than a parent replacement.
Click to expand...
Except that it's literally not "no different than googling" because it gets things wrong. So many studies have shown how wrong LLMs can be about basic information.

It's completely unlike Googling because it doesn't require any thinking on the part of the user, which discourages critical thinking about the answer you're given. When you Google something and then look through the different sites presented, you have to think about if those sites are reputable resources and then determine what's true based on what you read. AI—especially an audio-only AI response—takes away all of that and confidently declares information without any ability to verify the source of the info on your part. That's bad, because the vast majority of people aren't going to take the time to follow up and verify the information later.

30 years ago if you were to find an encyclopedia that had lots of info but some of it was straight up incorrect or presented in a way that wasn't accurate, you would not buy that encyclopedia and the company would be shamed for publishing incorrect information. Why aren't we similarly critical of AI?
 
tech_enthusiast_ said:
Except that it's literally not "no different than googling" because it gets things wrong. So many studies have shown how wrong LLMs can be about basic information.

It's completely unlike Googling because it doesn't require any thinking on the part of the user, which discourages critical thinking about the answer you're given. When you Google something and then look through the different sites presented, you have to think about if those sites are reputable resources and then determine what's true based on what you read. AI—especially an audio-only AI response—takes away all of that and confidently declares information without any ability to verify the source of the info on your part. That's bad, because the vast majority of people aren't going to take the time to follow up and verify the information later.

30 years ago if you were to find an encyclopedia that had lots of info but some of it was straight up incorrect or presented in a way that wasn't accurate, you would not buy that encyclopedia and the company would be shamed for publishing incorrect information. Why aren't we similarly critical of AI?
Click to expand...
Productive struggle is what you are talking about. It's not the answer, it's how you get to the answer, dealing with ambiguity and uncertainty.
 
  • Like
  • Love
Reactions: turbineseaplane, jlnr and tech_enthusiast_
tech_enthusiast_ said:
Except that it's literally not "no different than googling" because it gets things wrong. So many studies have shown how wrong LLMs can be about basic information.
Click to expand...
Web researching can also get things wrong. This is a case of throw the baby out with the bathwater. Because AI isnt 100% it’s useless. That’s not really so. I wouldn’t be researching sensitive topics cavalierly.
tech_enthusiast_ said:
It's completely unlike Googling because it doesn't require any thinking on the part of the user, which discourages critical thinking about the answer you're given.
Click to expand...
If you take the value of the web search at face value it’s exactly like using AI. Maybe you don’t and that is great but many will and do. There is no guarantee that web searching is completely correct either. Both AI and web searching have their place and time.
tech_enthusiast_ said:
When you Google something and then look through the different sites presented, you have to think about if those sites are reputable resources and then determine what's true based on what you read. AI—especially an audio-only AI response—takes away all of that and confidently declares information without any ability to verify the source of the info on your part. That's bad, because the vast majority of people aren't going to take the time to follow up and verify the information later.
Click to expand...
I feel anything you write about the difference in AI generated response vs a web search response is rationalization. One can accept wrong web searches as fact.
tech_enthusiast_ said:
30 years ago if you were to find an encyclopedia that had lots of info but some of it was straight up incorrect or presented in a way that wasn't accurate, you would not buy that encyclopedia and the company would be shamed for publishing incorrect information. Why aren't we similarly critical of AI?
Click to expand...
Encyclopedia companies were notorious for writing incomplete and biased articles.
 
  • Like
Reactions: BigMcGuire and germanbeer007
tech_enthusiast_ said:
Except that it's literally not "no different than googling" because it gets things wrong. So many studies have shown how wrong LLMs can be about basic information.

It's completely unlike Googling because it doesn't require any thinking on the part of the user, which discourages critical thinking about the answer you're given. When you Google something and then look through the different sites presented, you have to think about if those sites are reputable resources and then determine what's true based on what you read. AI—especially an audio-only AI response—takes away all of that and confidently declares information without any ability to verify the source of the info on your part. That's bad, because the vast majority of people aren't going to take the time to follow up and verify the information later.

30 years ago if you were to find an encyclopedia that had lots of info but some of it was straight up incorrect or presented in a way that wasn't accurate, you would not buy that encyclopedia and the company would be shamed for publishing incorrect information. Why aren't we similarly critical of AI?
Click to expand...
Well you did mention critical thinking skills so what a good place to use that and also learn to fact check AI. People only trust AI because they have not been told anything else, so teach them that as the world is moving that way. I certainly double check AI's work when thinks look wonky or out of place (use ChatGPT Pro). It is a tool but I still will double check and verify the really important stuff or if things look really out of place.

Why are we not similarly critical of AI? Because people like watching TikTok and Instagram videos over reading books and actually discussing issues and willing to have their ideas challenged. There is a generation of people who cannot handle opposition to their thoughts and opinions and are essentially incapable already of critical thinking outside of the echo chamber their algorithms feed them. It is a tool to use and has benefits and just like the internet with all its flaws if you learn how to use it properly it is a powerful tool.

You might have missed it but the first thing that comes up with google now is an AI response to your query. How long before people don't even scroll beyond that?
 
  • Like
Reactions: BigMcGuire
tech_enthusiast_ said:
No, where did I say anything about looking it up on the computer when you get home?

What I'm saying is that it's better to engage in a discussion about the topic and think critically about it than to just give instant gratification...particularly when you're currently giving your child answers from a source that is known to be wildly incorrect about things a lot of the time.
Click to expand...

You did say 'we can figure it out when we get home,' but fair enough if you meant something different. On the accuracy point though, the concern about AI being unreliable doesn't really apply to well-established science questions aimed at a five year old. And even if it somehow gets it wrong, it's not like he's going to remember. That's about as low-risk a use case as exists.

AI is a tool; one I find incredibly helpful and empowering and one I use daily. People have been decrying the latest technology as ruinous for learning since literally the dawn of civilization. Plato wrote that Socrates thought that writing things down was bad for learning because it would destroy memory and understanding.

When the internet went mainstream there were waves of concern about kids getting instant answers from search engines rather than learning to research, that it would destroy attention spans, that children would stop developing critical thinking because they could just Google everything. There were serious op-eds and academic papers arguing that Wikipedia was making students intellectually lazy and that the ability to find answers instantly was harmful. Same thing with calculators in the 70s and 80s too.

If we want to have a discussion about whether real-time, conversational, authoritative-sounding answers are qualitatively different from a search result that still requires the user to evaluate and synthesize then I think that's valid discussion we can have. But I don't think it's fair to decry it "depressing" and insinuate I'm doing my kid a disservice because I'm using a tool that I find incredibly helpful and my kid is going to learn about anyway the nanosecond they start using technology in his school.
 
  • Like
  • Haha
Reactions: eltoslightfoot, iMean, artifex and 1 other person
wonderings said:
I am curious, do you drive a Tesla? I have had one through work for about a year now and would not want to go back to CarPlay. The whole system is butter smooth and works really well, CarPlay would look incredibly out of place and childish in the Tesla eco system.
Click to expand...
Have a Model Y, and miss Carplay everytime I get in it. The apps for music suck and texting/messages are very subpar. Any break in cell service and the music apps completely forget what they are doing and never resume. Also, getting logged out of Apple music constantly is tiresome, cant even play downloaded music without just doing bluetooth. Its really annoying when I'm driving the car and my wife's messages are still coming through even on my profile.

Though I will say Grok has been entertaining for the kids when in story mode. Have not really tried it for anything else important.
 
decimortis said:
I ask this in all seriousness, but can someone tell me a true, real-world advantage to having a chatbot in my car with me? I watched the little Grok video in the article for a potential answer. My question still stands.
Click to expand...
two examples,
In Brazil McDonald's, for the World Cup, has a different sandwich every day of the week.
I asked Grok an it told me the sandwich of the day, and asked if I wanted to know the rest. then he listed all of them.

I forgot the name of a medication. On the way to the pharmacy I can talk with grok, until I remember it.

I never had to stop the car and now with the update, I won't even have to take my phone.

Both instances I tried to use ChatGPT voice for the CarPlay convenience, but it got both questions wrong.
 
  • Like
Reactions: AL2TEACH
You must log in or register to reply here.
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.
Top Bottom
Back