What's new
By:

Lucid AI Agent kKim Java YouTube video

Halodde said:
Did I hear correctly that this is coming to Air owners in the next month or so? And if so, is this the first feature that requires the upgraded chip?
Click to expand...
1775314529435.webp
 
I understand it is really early software and I'm not dissing it by any means.

BUT, when the video is sped up while waiting for the response, it is ingenuine on behalf of Kim, the creator of the video to represent what is happening. The video hits for her are clicking - a couple threads have already referred to it, so it is accomplishing what Kim wants, but I stopped watching. I want reality, not another person's decision on what they think I'l like the most. Happens in national news shows every single night. Share the information, but don't shape the information.

Let me know if it has anything I should know.

🧦
 
A lot of what the vehicle companies are demonstrating are what I can already accomplish by chatting with Gemini on my Pixel phone. So in that sense, not impressive. But also not surprising: few companies have the resources to develop their own foundational multi-modal models. X.ai is doing it, but hemorrhaging billions in the process. Same deal with companies like Open AI. Google is making a lot of money, although I'm not sure whether their AI investments are now paying for themselves, or being subsidized by their other lines of business.

That said, seamless integration is useful, and so bringing the functionality into the car's UI is nice. I don't yet own a Gravity, so I can't test it, but I do wonder whether AA linked to my Pixel phone would allow me to pretty seamlessly chat with Gemini while driving today.
 
TahoeMichael said:
A lot of what the vehicle companies are demonstrating are what I can already accomplish by chatting with Gemini on my Pixel phone. So in that sense, not impressive. But also not surprising: few companies have the resources to develop their own foundational multi-modal models. X.ai is doing it, but hemorrhaging billions in the process. Same deal with companies like Open AI. Google is making a lot of money, although I'm not sure whether their AI investments are now paying for themselves, or being subsidized by their other lines of business.

That said, seamless integration is useful, and so bringing the functionality into the car's UI is nice. I don't yet own a Gravity, so I can't test it, but I do wonder whether AA linked to my Pixel phone would allow me to pretty seamlessly chat with Gemini while driving today.
Click to expand...

We're supposed to be able to chat with Gemini using AA now, but for some reason haven't been able to use it - only Google Assistant.
 
Gotta chase that stock price I guess. I hope there's at least someone in a basement somewhere working on actual useful features/bug fixes.
 
knewman said:
Gotta chase that stock price I guess. I hope there's at least someone in a basement somewhere working on actual useful features/bug fixes.
Click to expand...
Google maps is my #1 want for Gravity. PAAK second.
 
TahoeMichael said:
A lot of what the vehicle companies are demonstrating are what I can already accomplish by chatting with Gemini on my Pixel phone. So in that sense, not impressive. But also not surprising: few companies have the resources to develop their own foundational multi-modal models. X.ai is doing it, but hemorrhaging billions in the process. Same deal with companies like Open AI. Google is making a lot of money, although I'm not sure whether their AI investments are now paying for themselves, or being subsidized by their other lines of business.

That said, seamless integration is useful, and so bringing the functionality into the car's UI is nice. I don't yet own a Gravity, so I can't test it, but I do wonder whether AA linked to my Pixel phone would allow me to pretty seamlessly chat with Gemini while driving today.
Click to expand...
Not all of us have phones with Gemini, so this would be a nice addition for those with no access to it. It's one of those things right? If it's useful, you use it, if not then you won't. As long as they can implement it without messing anything else up (I know, I know, so we'll see 😂 You'd expect they will reach the point where they can walk and chew gum at the same time) I'd like to have the choice to use the feature or not.
 
9point81 said:
Not all of us have phones with Gemini, so this would be a nice addition for those with no access to it. It's one of those things right? If it's useful, you use it, if not then you won't. As long as they can implement it without messing anything else up (I know, I know, so we'll see 😂 You'd expect they will reach the point where they can walk and chew gum at the same time) I'd like to have the choice to use the feature or not.
Click to expand...
It'll be interesting to see how they price it. The investor day implies that this kind of thing will be a subscription feature. Which makes sense, voice chat with an LLM burns a lot of CPU time on the back end. But if Lucid wants to charge me for an AI assistant subscription, I'd have to think twice. That'll be a monthly fee for an assistant I can only use in the car. The appeal to me of something like Gemini is that I can access it from every device I own. So an advantage is just one subscription price. But even more than that, all the data and history is unified, which improves the utility a lot.

Although Gemini is a Google product, they do make it available to the iPhone users via an app in the Apple app store. And if a person prefers an AI assistant from some other company, they pretty much all make them available across the commonly used device types.
 
TahoeMichael said:
It'll be interesting to see how they price it. The investor day implies that this kind of thing will be a subscription feature. Which makes sense, voice chat with an LLM burns a lot of CPU time on the back end. But if Lucid wants to charge me for an AI assistant subscription, I'd have to think twice. That'll be a monthly fee for an assistant I can only use in the car. The appeal to me of something like Gemini is that I can access it from every device I own. So an advantage is just one subscription price. But even more than that, all the data and history is unified, which improves the utility a lot.

Although Gemini is a Google product, they do make it available to the iPhone users via an app in the Apple app store. And if a person prefers an AI assistant from some other company, they pretty much all make them available across the commonly used device types.
Click to expand...
Yeah, I’m in your boat. I don’t really need AI when driving. Just good realtime navigation is sufficient.
 
It's actually Open AI's LLM that Lucid is implementing. Did I hear that correctly?
 
prgarg said:
It's actually Open AI's LLM that Lucid is implementing. Did I hear that correctly?
Click to expand...
I missed any mention of the LLM vendor, but I wouldn't be surprised if you're correct. Their LLM is quite good, and they're aggressively seeking licensees for it. As an auto company I would have some minor concern about stability and reliability of Open AI's LLM, but switching to a different provider also shouldn't be that hard if they ever decide to do that.
 
TahoeMichael said:
A lot of what the vehicle companies are demonstrating are what I can already accomplish by chatting with Gemini on my Pixel phone. So in that sense, not impressive. But also not surprising: few companies have the resources to develop their own foundational multi-modal models. X.ai is doing it, but hemorrhaging billions in the process. Same deal with companies like Open AI. Google is making a lot of money, although I'm not sure whether their AI investments are now paying for themselves, or being subsidized by their other lines of business.

That said, seamless integration is useful, and so bringing the functionality into the car's UI is nice. I don't yet own a Gravity, so I can't test it, but I do wonder whether AA linked to my Pixel phone would allow me to pretty seamlessly chat with Gemini while driving today.
Click to expand...
Yeah, I think that's where a lot of these demos feel a bit underwhelming. The AI itself isn't really the impressive part anymore, most of us can already do this on our phones.

The interesting part is when it can actually interact with the car instead of just answering questions. Like changing settings, knowing the battery state, planning a route around charging, controlling climate, etc. That's where a built-in system could have an advantage over just running Gemini through Android Auto.
 
Clithapping said:
...The interesting part is when it can actually interact with the car instead of just answering questions. Like changing settings, knowing the battery state, planning a route around charging, controlling climate, etc. That's where a built-in system could have an advantage over just running Gemini through Android Auto.
Click to expand...
Available now on a Volvo near you!
It works quite well and has the entire internet at its "fingertips".
 
Clithapping said:
Yeah, I think that's where a lot of these demos feel a bit underwhelming. The AI itself isn't really the impressive part anymore, most of us can already do this on our phones.

The interesting part is when it can actually interact with the car instead of just answering questions. Like changing settings, knowing the battery state, planning a route around charging, controlling climate, etc. That's where a built-in system could have an advantage over just running Gemini through Android Auto.




Speaking of connecting different apps and systems to make things actually interact and automate tasks, if you ever want to build custom integrations or smart agents without dealing with complex backend code, you can check out https://latenode.com/products/agent-builder.
Click to expand...
Just thought of one more thing. If the built-in AI can keep working even when your phone isn't connected or you have spotty reception, that could actually be a nice advantage too. It's one of those things you don't really think about until you need it.
 
I would trade all of this AI nonsense for genuine accessibility features.

Can I be read the available menu buttons and select menu buttons by voice? Could I request my unread text messages be read with the sender info and send back a response? Could the same be done for email, my calendar, etc?

The days I've been alone in the car thinking about a piece of code. Wouldn't it be great to sync up to my codebase with permissions, open the requested file, read the lines of code, and be able to modify by voice on a remote server I own? Compile, test, and run with the outputs reported back.

Now replicate that for anyone's day to day activities. Whatever they work with being available in a reasonable manner and modifiable.
 
You must log in or register to reply here.
Back
Top