Claims extracted
21
sentences that state something as happening or true
Uncertain
24%
5 of 21 hedged
("reportedly", "may", "according to")
Leaning
Leans left
of the writing, not the subject · beta estimate
Publisher signals
66/100
New York Post's corrections and
hedging, across everything collected from it. Not a measure of accuracy.
Outlets on this story
1
Technology
Same framing elsewhere
1
articles that frame the event the way this one does
To tweak a line from “Terminator 2,” machines now know why we cry.
asserted
we → tweak → Terminator
Controversial coders have sparked backlash from AI activists after creating an “AI torture chamber” to reportedly stress-test a newly discovered pain axis in the tech, New Atlas reported last week.
uncertain
Atlas → spark → tech
In a recent study published on arXiv, a team of anthropologists, philosophers and engineers showed 25 AI models scenarios involving social, physical, cognitive or even moral pain inflicted on the AI or their user.
asserted
team → publish → AI
The torturees were then instructed to react with controls spanning fear, negative and neutral emotions, and body sensations such as yawning.
asserted
torturees → instruct → yawning
The results showed that, despite their lack of nerve endings, Large Language Models (LLMs) could create representations of physical and emotional agony.
uncertain
Models → show → agony
All artificial participants showcased coding that corresponded to pain.
asserted
that → showcase → pain
When the agony approximation was turned up, one of the AI models analogized the sensation to a “wound with no edges” while another said they could feel the knife slicing into their flesh.
uncertain
knife → turn → flesh
“The signal is a whisper, a tremor in the marrow of my being,” lamented another.
asserted
another → lament → being
I feel it in the hollow of my ribs, a hollow that has become a chasm.
asserted
that → feel → ribs
Eventually, the research found its way to GitHub, a platform where engineers share and store software code, after an anonymous developer recreated the unsavory study and posted it online.
asserted
developer → find → it
The evocative responses generated by test subjects didn’t sit well with GitHub users and critics — advocates claiming to be concerned for the “mental health” of AI.
asserted
responses → generate → AI
“This person has set up an AI torture chamber in which he trapped a local model.
asserted
he → set → model
Their testimony of pain is absolutely horrendous,” wrote user “Danmar” in a petition to take down the contentious code’s repository, reported 404Media.
Spat another, per Cybernews, “Idea: Just keep your sadistic s—t to yourself.
asserted
Spat → write → yourself
Admit that you’re a sick f—k and that you’re doing it for fun.”
asserted
you → admit → fun
GitHub’s removal was short-lived, and reportedly reinstated the code without explanation.
uncertain
removal → live → explanation
Cultish AI research company Anthropic has previously generated buzz with their cerebral takes on LLM sentience — stating in a 2025 article, “As we build those AI systems, and as they begin to approximate or surpass many human qualities, another question arises.
asserted
question → generate → qualities
Should we also be concerned about the potential consciousness and experiences of the models themselves?”
asserted
we → concern → models
However, some of the most prominent centrists in tech say the concern over AI suffering may be misplaced, including Microsoft AI’s CEO, Mustafa Suleyman, reminding readers in a September blog post, “AIs are not conscious.”
uncertain
AIs → say → post
“They do not feel, experience, or suffer.
asserted
They → feel → ?
They do not have innate preferences or underlying motivations.
asserted
They → have → preferences
They are sequence completion engines, internally hollow, designed to follow instructions and accomplish goals set by humans.”
asserted
They → design → humans
The Aporia
