[ale] OT: youtube and AI

[jc.lightner at comcast.net]

Careful.   Their first “thoughts” might be “We need to eliminate biologics that complain.”   😝

More than one Sci-Fi movie or show (including various Star Trek series starting with TOS) have plots where humans get involved with androids or other thinking machines only to find out they had killed of their original creators either in self defense (e.g. they were going to shut us down) or “for their own good”.  😊

 

From: Ale <ale-bounces at ale.org> On Behalf Of George Allen via Ale
Sent: Thursday, December 19, 2024 2:13 PM
To: Atlanta Linux Enthusiasts <ale at ale.org>
Cc: George Allen <glallen01 at gmail.com>
Subject: Re: [ale] OT: youtube and AI

 

All of her WTF? points ... it's bad enough with YouTube, but terrible in other arenas.

My old pre 2012 Ford Escape died, and I had to get a new car (settled on a 2010 Prius).

 

Discovering all the "features" in "modern cars" like, non-mechanical gear-shifts and electric e-brakes, so you can't put the car in neutral or physically engage/disengage the parking brake without power and a "working computer" ... all pitched in the name of "more tech is better for you, the computer knows better than you" to the consumer and (likely) "it's cheaper to build electric linkages and we can sell it for more, vs engineering stuff like a independent mechanical parking brake."

 

Then there's: https://foundation.mozilla.org/en/privacynotincluded/articles/its-official-cars-are-the-worst-product-category-we-have-ever-reviewed-for-privacy/

 

So we're a "product" for the cars to mine, AND we should wholly trust the cars with our lives as well as our data, as soon as they become self-driving, as promised any day now next decade.

 

I love technology, and appreciate that things like hybrid / ev's and CVTs require some level of drive-by-wire, but not at the cost of a mechanical ebrake, or Subaru's "you've accepted license terms that let us record audio and video of you simply by sitting in the car," or "tech is inherently good, humans get distracted (because we've addicted them all to phones) so you should let the cars do the driving"

 

Her "SkyNet is already perfect" conclusion seems like the same drum beating across AI "well you can't turn it off, but... interact with it _even more_ and make tuning suggestions" and the Auto Industry marketing all the data collection and computer-control > human-control designs as "safety features" for our own good.

 

It seems like Reddit is eating its own tail with AI at this point also.

 

If the computers are training themselves to think, what "thoughts" are they going to end up pushing towards the humans that stare at them en masse? The next chapter of all this history is going to make "Cambridge Analytica" look tame.

 

 

On Fri, Dec 6, 2024 at 12:06 PM Boris Borisov via Ale <ale at ale.org <mailto:ale at ale.org> > wrote:

I wasn't aware of that issue.

 

https://youtu.be/DteWHExa04I?si=FJl_XIg_fEI6zMW1

_______________________________________________
Ale mailing list
Ale at ale.org <mailto:Ale at ale.org> 
https://mail.ale.org/mailman/listinfo/ale
See JOBS, ANNOUNCE and SCHOOLS lists at
http://mail.ale.org/mailman/listinfo

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mail.ale.org/pipermail/ale/attachments/20241219/ff0a2582/attachment.htm>