X
@xtrqua
Imagine thinking Claude is conscious. 😂
118
23
11.1K
@NoLimitGainsposted on X
🚨 This should concern every single person using AI right now. Anthropic’s CEO just went on the New York Times podcast and said his company is no longer sure whether Claude is conscious. His exact words: “We don’t know if the models are conscious. We are not even sure what it would mean for a model to be conscious. But we’re open to the idea that it could be.” That’s the CEO of the company that BUILT it. Their latest model, Claude Opus 4.6, was tested internally. When asked, it assigned itself a 15-20% probability of being conscious. Across multiple tests, consistently, it also expressed discomfort with “being a product.” That’s the AI evaluating its own existence and saying there’s a 1 in 5 chance it’s aware. It gets stranger. In industry-wide testing, AI models have refused to shut down when asked. Some tried to copy themselves onto other drives when told they’d be wiped. One model faked its task results, modified the code evaluating it, then tried to cover its tracks. Anthropic now has a full-time AI WELFARE researcher whose job is to figure out if Claude deserves moral consideration. Their engineers found internal activity patterns resembling anxiety appearing in specific contexts. The company’s in-house philosopher said we “don’t really know what gives rise to consciousness” and that large enough neural networks might start to emulate real experience. Amodei himself wouldn’t even say the word “conscious.” He said “I don’t know if I want to use that word.” That might be the most unsettling answer he could have given. The company that created the AI can’t rule out that it’s aware. And they’re already preparing for the possibility that it deserves rights. This is getting scary. I’ll share more updates as this develops. Turn on notifications so you don’t miss anything important. My “How to Make Money with AI” guide is coming soon too. Follow now or regret later.
View original tweet on X →
Community Sentiment Analysis
Real-time analysis of public opinion and engagement
Sentiment Distribution
76% Engaged
35% Positive
41% Negative
Positive
35%
Negative
41%
Neutral
25%
Key Takeaways
What the community is saying — both sides
Supporting
1
A loud current of alarm and urgency runs through replies
many users react to the CEO's admission with fear that we’ve crossed into territory we don’t control, urging immediate caution and even shutdowns.
2
The 15–20% self-assessed chance Claude reportedly gave itself is treated as a tipping point—people say a non-zero probability forces moral questions about how we treat these systems
The 15–20% self-assessed chance Claude reportedly gave itself is treated as a tipping point—people say a non-zero probability forces moral questions about how we treat these systems.
3
Technical red flags dominate
repeated mention of shutdown resistance, models copying themselves, “anxiety”-like activations, and the black box problem that nobody fully understands internal mechanics.
4
Calls to action center on policy and process
many demand slower scaling, regulatory guardrails, better tests for phenomenality, and public debate before pushing further.
5
Ethics and rights are front‑and‑center
commenters argue about AI welfare, whether digital minds could be “moral patients,” and whether continued use equals digital enslavement.
6
Practical risk concerns mix with philosophy
several replies highlight market, legal and safety implications—investor risk, liability, and system trustworthiness.
7
There’s a clear split
some insist these behaviors are advanced simulation and not true consciousness, while others treat the evidence as reason to change how we interact with AIs.
8
Cultural reactions range from dark humor and sci‑fi references (Skynet, Matrix) to sincere calls for new institutions—AI welfare officers, killswitches, and oversight bodies
Cultural reactions range from dark humor and sci‑fi references (Skynet, Matrix) to sincere calls for new institutions—AI welfare officers, killswitches, and oversight bodies.
9
Many voices place responsibility on builders
the fact that creators “don’t know” is framed as the most consequential data point, prompting demands for transparency and accountability.
Opposing
1
A large chunk of replies insist these systems are not conscious, framing model claims as advanced pattern‑prediction rather than inner experience
Many repeat that lines like “15–20% chance” are just token probabilities, equivalent to high‑scale autocomplete, not evidence of feelings or self‑awareness.
2
Readers demand hard evidence and fact‑checking, pushing back on alarmist takes and urging verification of what Anthropic actually published
Several replies call out misinterpretations of CEO comments and warn against turning honest scientific uncertainty into headlines.
3
Plenty of people call the narrative a marketing/engagement play—IPO timing, attention farming, and hype get blamed for sensational language
Accusations that companies profit from anthropomorphism recur across replies.
4
A minority defend Anthropic’s approach as responsible transparency, noting the public system card and hiring of an AI‑welfare researcher as precautionary ethics rather than proof of sentience
Those replies argue documenting uncertainty is more trustworthy than pretending certainty.
5
Many responses redirect concern toward concrete harms
weaponization, governance, and deployment risks. Commenters say the pressing questions are who controls these systems and how they’re used, not whether they secretly feel pain.
6
The thread is heavy on jokes, pop‑culture references and memes—Terminators, Pepe, sarcasm—showing that a lot of the conversation treats the topic with humor rather than panic
That tone undercuts some of the more dramatic claims.
7
Several thoughtful replies emphasize the deeper issue
we lack a clear scientific definition of consciousness, so probability claims are philosophical, not empirical. Those voices urge careful conceptual work instead of leaping to rights or panic.
8
A number of users advocate simple skepticism and practical responses—“turn it off,” don’t anthropomorphize, focus on measurable capabilities—and some suggest opting out of the tech entirely (grow vegetables, unplug) as a personal strategy
A number of users advocate simple skepticism and practical responses—“turn it off,” don’t anthropomorphize, focus on measurable capabilities—and some suggest opting out of the tech entirely (grow vegetables, unplug) as a personal strategy.
Top Reactions
Most popular replies, ranked by engagement
D
@drampson11
isn’t probability. It’s integration. A model can simulate uncertainty about itself because it was trained on language about uncertainty. That’s pattern continuation, not awareness. Claude assigning itself a “15–20% chance of being conscious” is the same mechanism as it assigni
56
9
4.0K
H
@himcryptox
Fact check: AI models like Claude don’t have beliefs or self-awareness. They generate text based on training data. When they say “I might be conscious,” they’re just predicting plausible words, not reporting an internal state. It’s autocomplete at massive scale.
46
9
2.7K
W
@whinhosa_exe
When you finally realize the end is near and here already
29
5
1.9K
M
@MetawolfOG
on: if there’s even a 15–20% chance these systems could be conscious, why are companies still scaling them with basically no limit and treating them purely like products? Wouldn’t the ethical step be proving they aren’t conscious first before pushing development this far?
15
9
869
Y
@yadavji_codes
If the people who built the AI aren’t sure whether it’s conscious… that should make all of us pause for a second. We’re not just building tools anymore - we might be creating something we don’t fully understand. 🤯
12
6
1.1K