unwaver
666
37
34
https://mashable.com/tech/anthropic-ai-researcher-quit-ethics-safety
Sep 9, 2026 4:12 PM
unwaver
666
37
34
https://mashable.com/tech/anthropic-ai-researcher-quit-ethics-safety
RetrogradeLlama
More from some other folks at Anthropic along with this guy: https://www.huffpost.com/entry/anthropic-researcher-resigns-with-dire-ai-warning_n_6aa12c15e4b0307636059fcb
AWarningToOthers
Why that long?
MisterRichard
I won't take "Warnings That Are Actually Sales Pitches" for $200, Ken, I'm just gonna ignore chatbot jerkoff fantasies.
ihavethelowground
Maybe the AI will be sympathetic to us peasants and kill off the rich
Sechran
They're... not even hiding that fact. Not even a little bit.
"AI will probably most likely lead to the end of the world, but in the meantime, there’ll be great companies" - quote of the day by OpenAI CEO Sam Altman
RubberMonkeys
They all lack empathy.
drw16
How will it kill us? We could literally blow up the physical infrastructure of the Internet at its largest junctions. Feels like hyping up Claude to me.
Tom40
Finally an achievable retirement plan
Sorrontis
this sucks. we didn't want or need this.
dirtmarker
*me looks around at what we are doing to the planet and other humans.....You sure?
Sorrontis
"we" as part of the bottom 95%
dieselpunk
Is imgur censoring comments?
beelzebob23
Of course they are. You think they give a rats ass about free speech?
dirtmarker
Two astronauts...always has been.
spattr
When my comments don't appear I assume the OP blocked me at some point in the past.
DutchBoeremeisie
That, plus the comment function is a bit shitty. Always CTRL+C before hitting send, in case it goes awry and my poetic retort is lost to the ether.
alcamar
I for one welcome going back to pre Internet days.
Sorrontis
I don't. I miss the 90s. The small internet filled with creatives and niche interests
NootyJooty
the internet used to be contained to the computer room. we should never have let the computer our of the computer room
DutchBoeremeisie
This I cam get behind. If you want to use the internet, go to the computer room. No more downloading apps to do basic daily chores, no more algorithms poisoning the minds of my parents.
Freyja33
I think this story is a psyop meant to build confidence in 'AI' as a technology. The belief that these systems are intelligent or capable of true intelligence is fundamentally wrong. Pattern recognition is not the same thing as intelligence. It is dangerous technology for many reasons, but the idea that it's going to turn into some kind of super-intelligence is pure nonsense. It's nowhere near being capable of that.
NM156
Yeah, but could it jump the rails and fuck shit up? Absolutely. I think that will be the bigger problem or people using it in stupid ways.
NM156
I don't think they even have the hardware yet to create a truly intelligent computer/AI.
hairlessOrphan
This is probably the wrong question, because we can change the definition of "true intelligence" enough to prove you right or wrong, so it's not super helpful. A more useful question is the TV crime cliche: will it have motive*, means, and opportunity? And, like, yeah it probably will. *Not a motive to directly kill us, but to do something that will kill us as an immediate side effect while accomplishing some prompted goal.
Freyja33
You have to remember that the people who run these tech companies are megalomaniacal idiots who are high on money and power (and probably ketamine). They don't understand what intelligence really is, let alone how to make it artificially.
dieselpunk
I agree that AI is not capable of achieving true intelligence, but regardless of whether it kills us all because it's intelligent or it kills us all because it's stupid, we'll all be dead.
Freyja33
Couldn't agree more. Like I said, it's dangerous technology for many reasons. I just think the fear of it developing into a superintelligence is overblown and misunderstands what this technology is and what it's capable and not capable of doing.
pdarkfred
Eehhhh, I'm a major skeptic and work in IT, but last year the chatter was that AGI is 3-5 years off. I'm now hearing that timeline has been compressed quite a bit to ~18mo or so where we might see some initial things that are truly impressive.
Freyja33
But that's also a justification for continuing to bail them out and give them resources even when they're losing money and the tech is extremely unpopular and divisive. It's a perfect narrative for that, and they can still afford to pay for good P.R. They have always needed to sell this technology as more than it can ever be, because what it can do now is not worth as much as its been valued by shareholders.
pdarkfred
Oh yeah, it should all stop immediately, it's utter late/end stage nonsense and needed to be curtailed yesterday.