2025.26 Update with Grok + Dashcam viewer

[TeslaKen]

Apple TV

Sponsored

 

[hemiarch]

Yes. Sorry about that. Still awesome though

 

[vandytom]

I’m so happy that Foundation is on again!!!

Please respect and enjoy the Emperor’s peace.

I almost thought you were talking about a relaunch of the Foundation cybertruck ?

 

[vandytom]

As a fan of his books, especially the Foundation Trilogy, there is a show?

Where TF you been? There’s like 2 seasons on Apple TV+ now ?

 

[FL370]

Where TF you been? There’s like 2 seasons on Apple TV+ now ?

Lol Apple TV is the only service we don't have.

 

[Gaximus]

All can be accomplished without interjecting politics into the AI. For those of you who read and appreciate Asimov, human politics should never factor into the logic system.
Simple rules.
Three specifically.

Not really true, AI is trained on data sets. The larger the data set the better the AI. The internet is the largest data set. and contains politics, in everything. I mean everything. Watch a instagram or YouTube of someone getting hit in the head, and the comments will quickly be like “clearly a trump voter” or “that’s what they get for voting for Biden”. AIs train on this. You can’t filter out parts of data and leave other data in. That becomes “who gets to decide?” Which then becomes political.

 

[Loudness]

AI will not ever be in any car or on any device that I own.

Boy, do I have some bad news for you about FSD...

 

[dalton108]

Boy, do I have some bad news for you about FSD...

Nope. Get caught up please.

 

[Loudness]

Nope. Get caught up please.

Whoops, I missed that your post was altered by Admin. Carry on.

 

[hemiarch]

Not really true, AI is trained on data sets. The larger the data set the better the AI. The internet is the largest data set. and contains politics, in everything. I mean everything. Watch a instagram or YouTube of someone getting hit in the head, and the comments will quickly be like “clearly a trump voter” or “that’s what they get for voting for Biden”. AIs train on this. You can’t filter out parts of data and leave other data in. That becomes “who gets to decide?” Which then becomes political.

Fair point

 

[riggster]

Also, correct. Also, we will not be allowed to debate the issue here. So this will get shut down and that post with the video WILL BE DELETED.

I’m trying hard to thread the needle and play by the rules so I haven’t ever offered any discussion about the substance of this issue because it definitely affects my usage of my CB specifically and all my Teslas in general. Here’s what I’ve consistently said since day one:

1. I won’t use, or share any data with a >>purposefully<< biased AI of any stripe. I won’t willingly agree to any TOC that gives my data, thoughts, political opinions or any other metric to an AI with a fucking agenda!

??That people don’t understand how insane that is stupefying to me.

2. My very sane concerns about this “anti-woke” agenda have been repeatedly validated in glorious public debacle again and again. It leads to Ultron, not Vison — 100%.

3. No thanks! Not in my truck and that makes it a legitimate forum topic if not a topic Admin is willing to allow to create massive flame wars and discord.

That’s why I kept it to: I don’t want this, I want to be able to opt out. I didn’t make this political Voldemort and TechnoKing did.

I said MH, admin censored it and the instant “hilarity” ensued. That’s pretty much a full rundown.

I think you may be the one who has a misunderstanding of how Grok, Claude, ChatGPT, and other LLMs are trained. I'm nearly finished with a master's in data science with a specialization in AI from Northwestern University and may overestimate or underestimate the general population's understanding of LLMs. Grok is trained in a static environment based on a set of data used for training at the time. They may use some social media, such as this forum or X (formerly Twitter), to train models. Still, conversations with the LLM aren't typically used for training by default for most providers—though for Grok specifically, xAI states that they may utilize user interactions, inputs, and results with Grok (e.g., prompts and responses) to train and improve the model unless you explicitly opt out via your X account settings. This is outlined in xAI's privacy policy. Similar opt-out options exist for tools like ChatGPT to prevent your chats from being used in training data. If your concern is about data privacy or bias, opting out is a straightforward way to address it without forgoing the tool entirely.

 

[REM]

For those who want to keep up with what Elon ACTUALLY says without the crazy MSM spin I'd recommend Jacob Hinton's daily channel "Everything Elon Musk said today". The videos are generally about 10 minutes daily. Getting information from the horses mouth without the political spin is perhaps a useful idea.

https://www.youtube.com/@jcubhilton/videos

It's surprisingly a great, to the point channel that packs in a ton of info!

 

[dalton108]

I think you may be the one who has a misunderstanding of how Grok, Claude, ChatGPT, and other LLMs are trained. I'm nearly finished with a master's in data science with a specialization in AI from Northwestern University and may overestimate or underestimate the general population's understanding of LLMs. Grok is trained in a static environment based on a set of data used for training at the time. They may use some social media, such as this forum or X (formerly Twitter), to train models. Still, conversations with the LLM aren't typically used for training by default for most providers—though for Grok specifically, xAI states that they may utilize user interactions, inputs, and results with Grok (e.g., prompts and responses) to train and improve the model unless you explicitly opt out via your X account settings. This is outlined in xAI's privacy policy. Similar opt-out options exist for tools like ChatGPT to prevent your chats from being used in training data. If your concern is about data privacy or bias, opting out is a straightforward way to address it without forgoing the tool entirely.

I currently hold a provisional AI patent. I do not now, nor will I ever trust xAI. If X becomes a financial service corp., I’ll never intentionally do any financial business with them. Others will do as they will.

Ignorance of how LLMs are trained/work is not my issue. I AM ignorant as to the terms of xAI’s TOS, and will remain ever thus due to my aforementioned lack of interest in their offerings.

How this works is: I ALONE determine the nature and extent of my engagement with any individuals or entities.

My engagement with Tesla is limited to my previous interest in obtaining and now owning and operating electric vehicles they’ve produced which suit my needs (and some stocks that I hold).

I use several LLMs that I have vetted for my personal and commercial use. None are produced by xAI; nor shall they ever be.

Thanks.

 

[1Mathieu1]

There better be an option to opt out of full integration. The last thing I want is to have it replace my current voice commands.

AI will not ever be in any car or on any device that I own. So, that leaves 2 possibilities for my Teslas: (1) Opt out, or (2) bye-bye.

I guess staying on 2025.20.6 indefinitely is a third option, but that’s not tenable, so, I sure better see a fucking toggle. ?

I need ASS, dumb summon, reliable FSD, a better charging curve, HomeLink, functional auto-wipe, and a remapping of the suspension over bumps + rear wheel drive bias— all waaaaay before I need this culture-warrior AI trash ? in my CB! ?

Goodbye ??... Hello Grok! ? ?

 

[dalton108]

Goodbye ??... Hello Grok! ? ?

Hard time following the bouncing ball, huh? Unfortunate, since there was one bounce. It’s opt in, so my Teslas will remain in my possession.

You, on the other hand … <click>.

Sponsored