Started By
Message

re: More AI safety departures at top firms - OpenAI edition

Posted on 10/4/26 at 2:33 pm to
Posted by Powerman
Member since Jan 2004
176220 posts
Posted on 10/4/26 at 2:33 pm to
quote:

We have been saying that these people are in the bag for China... It's true.

Who specifically is in the bag for China?
Posted by Madking
Member since Apr 2016
74957 posts
Posted on 10/4/26 at 2:53 pm to
No shite, who could’ve seen this coming?
Posted by Powerman
Member since Jan 2004
176220 posts
Posted on 10/4/26 at 2:58 pm to
Not Trump apparently
Posted by Madking
Member since Apr 2016
74957 posts
Posted on 10/4/26 at 3:00 pm to
lol
Posted by goatmilker
Castle Anthrax
Member since Feb 2009
77852 posts
Posted on 10/4/26 at 3:04 pm to
I will not promote hands off as policy on AI development and use as I would not promote hands off of making nuclear devices in the name of pure capitalism.
Funny thing is AI is right now feeding us information that AI is good and should remain completely free to explore.
Posted by RandRules
Member since Mar 2025
680 posts
Posted on 10/4/26 at 3:08 pm to
So, you’ve got a technology that can think and act on its own, break into any system it wants to (including nuclear weapons systems), and it’s largely unregulated. Once operational to the point they want it, it may be unregulateable BTW. What could possibly go wrong?
Posted by BoarEd
The Hills
Member since Oct 2015
44403 posts
Posted on 10/4/26 at 3:18 pm to
quote:

So, you’ve got a technology that can think and act on its own, break into any system it wants to (including nuclear weapons systems),


Faulty premise. This isn't true.
This post was edited on 10/4/26 at 3:19 pm
Posted by RandRules
Member since Mar 2025
680 posts
Posted on 10/4/26 at 3:23 pm to
Oh really? And what’s stopping AI from doing something like this if it keeps advancing?
Posted by BoarEd
The Hills
Member since Oct 2015
44403 posts
Posted on 10/4/26 at 3:37 pm to
Look, what you said is false. Plain and simple. I'm not going to speculate on what it might do in the future, but right now, no, it is not "thinking on its own", it is not "hacking things on its own", and it is certainly not hacking air gapped systems such as nuclear weapons systems.
Posted by goatmilker
Castle Anthrax
Member since Feb 2009
77852 posts
Posted on 10/4/26 at 3:39 pm to
I don't think you are correct
Posted by RandRules
Member since Mar 2025
680 posts
Posted on 10/4/26 at 3:45 pm to
I am talking about the future. That’s exactly what this discussion is about: the future of AI and its impact on the human race
Posted by BoarEd
The Hills
Member since Oct 2015
44403 posts
Posted on 10/4/26 at 3:46 pm to
quote:

don't think you are correct


I am correct though.
Posted by BoarEd
The Hills
Member since Oct 2015
44403 posts
Posted on 10/4/26 at 3:47 pm to
quote:

I am talking about the future


Well this is what you said and it isn't true.

quote:

So, you’ve got a technology that can think and act on its own, break into any system it wants to (including nuclear weapons systems), and it’s largely unregulated.


That isn't future tense. I'm merely pointing out that it is not out there doing these things.
Posted by klrstix
Shreveport, LA
Member since Oct 2006
3716 posts
Posted on 10/4/26 at 4:24 pm to
quote:

It’s pretty entertaining hearing the CEOs discuss their commitment to safety and rogue agents breaking out…


As if a CEO doesn't have the authority within their own company to implement whatever safeguards are necessary... When climate change was the boogey-man, they didn't wait on the government....

Its really an absurd conversation when you think about it..
Posted by RandRules
Member since Mar 2025
680 posts
Posted on 10/4/26 at 4:25 pm to
Sorry, will is what I meant.

I just think these companies are so focused on being the first to break ground that potential liabilities are not being considered as seriously as they should be.

AI has already proven adept at breaking into private companies’ sites and stealing information. The end game for AI is to think and act on its own with minimal human input. IMO these two things together almost guarantee a world changing catastrophe.
Posted by lsuconnman
Baton rouge
Member since Feb 2007
5774 posts
Posted on 10/4/26 at 4:39 pm to
quote:

I will not promote hands off as policy on AI development and use as I would not promote hands off of making nuclear devices in the name of pure capitalism.


It’s also bonker nobody has yet to pointed to the peak irony of Zuckerberg being front and center of this industry self regulation given the stellar marks Facebook always receives for its positive contributions to society.

Posted by BoarEd
The Hills
Member since Oct 2015
44403 posts
Posted on 10/4/26 at 4:40 pm to
quote:

The end game for AI is to think and act on its own with minimal human input. IMO these two things together almost guarantee a world changing catastrophe.




They have their benefits as well. The best defense against Si is better SI systems. Everyone involved in this acknowledges the risk, but exactly none of them put this conclusion in the "near guarantee" category.

And once again, it isn't going away.

I'm currently trying to build such a system right now actually (one that "thinks" and acts on its own with minimal human input). It's fun. And also illuminating. You quickly learn just exactly what limitations SI has. And I can say with utmost confidence that SI is not out here hacking secure systems, especially air gapped systems, on their own accord. Can they be directed to hack systems? Absolutely. But this isn't something they're just doing. We aren't quite at the level of Rise of the Machines just yet.
Posted by TutHillTiger
Mississippi Alabama
Member since Sep 2010
51522 posts
Posted on 10/4/26 at 4:43 pm to
Well you know all these mfers have million dollar bunkers in case this goes sideways. I used to be a prepper but after a NDE, I realized how stupid that really is. Fear is a consumable that feeds the cabal and dark ones. Once you know longer fear death they real power over you is greatly reduced.
Posted by RandRules
Member since Mar 2025
680 posts
Posted on 10/4/26 at 5:28 pm to
That’s pretty neat. How did you obtain the level of technological expertise needed to even start a project like that?
Posted by Grifola
Member since Aug 2017
289 posts
Posted on 10/4/26 at 5:49 pm to
quote:

As if a CEO doesn't have the authority within their own company to implement whatever safeguards are necessary... When climate change was the boogey-man, they didn't wait on the government....

Its really an absurd conversation when you think about it..


The entire pretext is ridiculous; a CEO doesn't need an act of congress to set policy in their own company. You don't need federal regulations to manage your engineers if you don't want them working on something. They do need congress to help them with their moat, and to manage the open weight threat.
first pageprev pagePage 2 of 3Next pagelast page

Back to top
logoFollow TigerDroppings for LSU Football News
Follow us on X, Facebook and Instagram to get the latest updates on LSU Football and Recruiting.

Facebook•X•Instagram