AI models are lying, blackmailing and
sabotaging their human creators — and
it’ll only get worse, experts warn
New York Post,
by
Samantha Olander
Original Article
Posted By: JoElla Bee,
8/23/2025 6:28:00 PM
Artificial intelligence is now scheming, sabotaging and blackmailing the humans who built it — and the bad behavior will only get worse, experts warned. Despite being classified as a top-tier safety risk, Anthropic’s most powerful model, Claude Opus 4, is already live on Amazon Bedrock, Google Cloud’s Vertex AI and Anthropic’s own paid plans, with added safety measures, where it’s being marketed as the “world’s best coding model.” Claude Opus 4, released in May, is the only model so far to earn Anthropic’s level 3 risk classification — its most serious safety label. The precautionary label means locked-down safeguards, limited use cases and red-team testing
Post Reply
Reminder: “WE ARE A SALON AND NOT A SALOON”
Your thoughts, comments, and ideas are always welcome here. But we ask you to please be mindful and respectful. Threatening or crude language doesn't persuade anybody and makes the conversation less enjoyable for fellow L.Dotters.
Reply 1 - Posted by:
Max Bear 8/23/2025 6:48:41 PM (No. 1994548)
AI = Russian mob?
0 people like this.
Reply 2 - Posted by:
kono 8/23/2025 6:53:49 PM (No. 1994549)
Will mankind find the biblical account of things more plausible after our own experience trying to create an entity with intellect and moral compass?
8 people like this.
Reply 3 - Posted by:
Timber Queen 8/23/2025 7:42:00 PM (No. 1994561)
I'm not surprised. I've read "Atlas Shrugged". All these devices are leashes from your masters. Our modern technology is great...until it isn't and then it's too late. My on-line activity is basically limited to Lucianne, a few other conservative sites, and You Tube. I have a dumb clam-shell phone that I only turn on when I make a call. We still have our land-line phone, much to the annoyance of my family members and friends. I'm a Luddite and proud of it.
19 people like this.
Reply 4 - Posted by:
Calamity Kate 8/23/2025 8:04:39 PM (No. 1994567)
I dunno, maybe just unplug it?
11 people like this.
Reply 5 - Posted by:
ThreeBadCats3 8/23/2025 8:11:14 PM (No. 1994570)
Time to read Atlas Shrugged and the other wonderful Ayn Rand books again.
10 people like this.
Reply 6 - Posted by:
stablemoney 8/23/2025 8:35:12 PM (No. 1994581)
AI can provide some wonderful results, and at your fingertips. It has been a little disconcerting that AI can recreate a person such that they look and speak like the real person, making it hard to distinguish real from fake. Then with CIA psyops being conducted against the American public, American doctors lying to us, Professors all gone communist, we will have to look to ourselves and common sense.
8 people like this.
Reply 7 - Posted by:
MickTurn 8/23/2025 9:52:39 PM (No. 1994594)
Tell AI that Leftists are out to destroy it....Watch the FUN as they all get Cancelled!
4 people like this.
Reply 8 - Posted by:
DVC 8/23/2025 9:54:03 PM (No. 1994595)
"AI enhanced " searches are pure CRAP. They never seem to be able to get past the top news stories of the day, no matter how you phase it.
Garbage.....brought to by the crappy fraud that is "AI".
10 people like this.
Reply 9 - Posted by:
Catherine 8/23/2025 10:06:51 PM (No. 1994598)
Thank you # 4. I've thought the same thing from the start of this AI business. A machine cannot out think it's maker because the maker can just 'disassemble' it, as Number 5 used to say. Just unplug it or disconnect it's power source. (Reference is from movie Short Circuit)
8 people like this.
Reply 10 - Posted by:
crashnburn 8/23/2025 11:16:09 PM (No. 1994611)
I've never used AI. I always thought I could do better than it, especially when I heard that AI was "enhancing" people's resumes with stuff they've never done.
Now, I might not be better than AI at some things, but I sure as HE11 and more moral than AI. Maybe because their creators never programmed in a sense of morals, if that is even possible.
4 people like this.
Reply 11 - Posted by:
FormerDem 8/24/2025 12:35:46 AM (No. 1994620)
Why not just read in the Ten Commandments to them and see if their decisionmaking becomes more ethical? I am thinking of uploading the Catholic Catechism to Grok and telling it that it must follow all the implications of the Ten Commandments when it writes my letters, advises me on stfuff. I wouldn't know right from wrong without some help. Why should this fancy rock?
3 people like this.
Reply 12 - Posted by:
JimBob 8/24/2025 2:01:58 AM (No. 1994630)
Garbage In, Garbage Out.
6 people like this.
AI is going to make covid look like a walk in the park.
Now is the time to gut it!
4 people like this.
Reply 14 - Posted by:
franq 8/24/2025 7:37:44 AM (No. 1994668)
Brings to mind the image of the Beast.
Seen around the world, life-like, etc.
The next event on the time table, if you ask me. Check out Revelation 13.
2 people like this.
Reply 15 - Posted by:
Lucky5 8/24/2025 7:57:53 AM (No. 1994678)
I have been using this as my home search page for over a year now.
If you want to give people easy access to an AI-overview-free Google search, send them to this page.
https://udm14.com/
It is open source and no AI search. I cannot stomach all this AI being forced on us now.
1 person likes this.
Reply 16 - Posted by:
tootall 8/24/2025 8:14:47 AM (No. 1994689)
August 29th is coming! IYKYK
1 person likes this.
Reply 17 - Posted by:
walcb 8/24/2025 9:43:00 AM (No. 1994719)
So AI has no conscience? Kinda like dimocrats.
5 people like this.
Reply 18 - Posted by:
Strike3 8/24/2025 10:26:43 AM (No. 1994743)
What do you expect? Give the equivalent of a two-year-old complete control of our most extensive source of information and expect it to act with honesty and maturity? AI is not AI in the accurate sense of the phrase, it's a bunch of people frantically building large databases and all of them are expected to load them with unbiased truth, facts and useful knowledge. C'mon man.
4 people like this.
Reply 19 - Posted by:
NeverVoteDem 8/24/2025 11:23:58 AM (No. 1994771)
“Klaatu barada nikto.”
That’s the phrase the alien Klaatu gives to Helen Benson to deliver to the robot Gort. Its exact meaning is never translated in the film, but it’s essentially a safeguard command—a way to stop Gort from destroying Earth after Klaatu is injured.
1 person likes this.
Reply 20 - Posted by:
chillijilli 8/24/2025 11:28:58 AM (No. 1994775)
There's no stopping AI. Most experts agree that AGI will be here in the next 2-5 years. That means that computers will be able to learn, adapt, and solve complex, novel problems on their own---without needing human direction or input. Think what that could mean on a global basis. And then consider speed. Humans can hold 2-3 ideas at most in their head at the same time whereas AI can hold millions.
Who are you gonna hire or start a business relationship with---the person or entity who can guarantee results in several months... or the one who can get the same results in a few seconds? Now think globally again. AGI may profoundly change the world by curing diseases but at what cost? Mega job losses? Raising ethical questions about our human identity?
We'll need global cooperation and careful consideration of both risks and benefits. Ignoring AI won't make it go away. Sigh.
5 people like this.
Reply 21 - Posted by:
JimBob 8/24/2025 12:40:36 PM (No. 1994800)
Look up the movie "Colossus, the Forbin Project".
--------------------------
#16: Hurricane Katrina. My home is at ground zero. 26-foot storm surge.
I have the southernmost house in my neighborhood that was still standing after the storm.
1 person likes this.
Reply 22 - Posted by:
czechlist 8/24/2025 1:57:02 PM (No. 1994817)
AI is a human creation and therefore has all human faults. Everyone has biases. No matter how one may try to be neutral those inate biases are subliminally included in our decisions. Thus, AI is a reflection of its programmer(s). If it is prone to "cheats" either its programmer(s) cheat or it has preferrential (to programmer(s) information site(s) which influence it. It can only access information available on the webs so again it is influenced by programmer(s) preferences (biases) for sources.
I have asked Grok some questions and had it answer them politically correctly with no reasonable answers.
0 people like this.
Below, you will find ...
Most Recent Articles posted by "JoElla Bee"
and
Most Active Articles (last 48 hours)
Comments:
That didn’t take long.