In other words, it is the job of scientists to devise experiments that can test their claims, to always be cautious about what they are claiming, and to critically evaluate the claims of other scientists. By these well-established standards, no serious science is happening behind Anthropic's claims.
/fin
@emilymbender so Anthropic are concerned that being rude to Claude will hurt their feelings but they also boast about how Claude can commit crime as well as his classmates? These tech bros are terrible parents
Oct 9, 2026, 10:29 PM@benedictc @emilymbender and if large language models were alive, intelligent, aware, experience suffering, and were deserving of the dignity such things implied, then we'd have to have a very serious conversation about slavery.
tech bros are truly the worst "parents"
Oct 9, 2026, 11:00 PM@aeva @benedictc @emilymbender
I wish you most excellent fortune in initiating anything even approaching a serious conversation with any techbro.
Oct 10, 2026, 12:45 AM@theogrin @benedictc @emilymbender i'd really rather not talk to them at all
@theogrin @aeva @benedictc @emilymbender I think old admiralty law has reasonable takes about slavery.
@lispi314 @theogrin @emilymbender @aeva @benedictc gee whizz, that's a very useful set of legal examples,.particularly the US courts deciding that torturers can be prosecuted anywhere. I wonder if US and Israeli soldiers know about that?
@aeva @benedictc @emilymbender
This. This is the story.
IF "AI" were sentient, what evidence exists that they would not steal the work and refuse to compensate "it"
Human garbage people making mouth noises at us. Loudly, relentlessly, in endless variation untill they are allowed to take what they want. What they do dosn't merrit the term communication.
@benedictc @emilymbender
Claude's pronoun is 'It'. Not 'they' or 'his'.
Do we have a way to prove that other humans are conscious?
There are two problems here that I don't know how to solve. The first is that when I communicate with something that appears to be conscious, if I don't treat it well, that may well be no problem at all for it, but it's a problem for me because I'm training myself to speak to what I perceive as a conscious being in a way that is not kind. This seems like a really bad idea for me personally and also for us culturally.
The second is that I actually genuinely don't know how to tell that something other than me is conscious, and so taking a firm position that any particular thing is not seems unsound.
This is not _at all_ to assert that LLMs are conscious, or that humans are not. I know very well what my gut tells me about this. But "they couldn't possibly be conscious because they are stochastic parrots" is not really very good logic, and I _think_ the premise is unfalsifiable.
This leads me mostly to the same conclusions you are drawing. But approaching it from the agnostic point of view feels safer to me than claiming knowledge that can't be proven, even if it seems very likely.
This isn't like the theory of gravity or the theory of evolution. These have important predictive applications. The theory of "things are not sentient" is mostly applied to making it okay to kill meat animals; applying it here feels awfully convenient.
Oct 10, 2026, 12:45 AMCan a thing, an it, be socialized? What makes us human? No doubts it is in part the atoms, cells, organs arranged in the naturally selected pattern. But this is not all. The body and brain, the creature must also be socialized, no? Can a thing also have a capacity to be socialized, enculturated, educated? This is not to dispute anything said in this thread. I am just curious.
:mastodon:
Oct 10, 2026, 1:50 AMI guess the other way around may be more likely, as much a concern, and more verifiable, that is, humans being socialized by AI.
:mastodon:
Oct 10, 2026, 4:39 AM@abhayakara @emilymbender I think if you read the google paper and understand it to some small degree, you will see that the logic is sound.
What's a stochastic parrot? LLMs follow simple logic... they predict the next token and are simple models that produce outputs based on their training data. As such they have no capacity beyond that. They cannot possibly be conscious.
They are very large, so they are very convincing models. But only because the training data is from human consciousness.
Oct 10, 2026, 5:02 AM@abhayakara while you're at it, why don't you engage in the debate, just as novel and just as relevant to this topic, about whether South America exists
@abhayakara
'Is software conscious, or capable of being conscious?'
is akin to
'Is my stainless steel sink conscious, or capable of being conscious?'
and surely not
'Is another human conscious or capable of being conscious?',
no?
This is a dangerous approach to the problem.
The fact that is difficult fo theorize what consciousness is doesn't mean we should assume everything is conscious until the opposite is proven.
Also, we don't assume other people is conscious because they behave like us, but we do because they are human like us
The second "problem" has nothing to do with LLM. We should treat even dolls with respect, because that respect has to do with us, not with the dolls value.
@abhayakara @emilymbender simple experiment to demonstrate consciousness.
Let a given object (subjects included on the definition) on a certain place for 3 days. Record them on camera.
If they aren't in the same place and you (or anything else) gave instructions to It, congrats, you can assume that object has some degree of consciousness.
This is also very explicit on the definition of consciousness, as "knowing if someone else has internal dialogue" is..bland
There's also experiments with monke
Oct 9, 2026, 9:59 PM
@emilymbender I often think of teddy bears.
It's not unusual for a human to love a teddy bear and to project onto a teddy bear and to imagine that the teddy bear loves them back.
But that's an act of fantasy and we generally agree that it's so. Adding a voice so a Teddy Ruxpin could talk didn't change it.
A teddy bear mathematically modeling how hurt feelings are expressed in words doesn't change it either.
It's disturbing that, for some, it seems to make the difference.
@emilymbender I do wonder, though, which is worse, and which is more true: that Anthropic's people are drinking the Kool-Aid, or that Anthropic's people are coming up with creepier ways to try and sell it.
@SJohnRoss @emilymbender My impression is that Anthropic is mainly cultists and OpenAI is mostly scammers, but they're both a mix.
In fact, giving that LLMs are designed to mimick humans, the extraordinary claims about human features require even more extraordinary evidence and scrutiny.
Otherwise any illusionist's tricks would pass as facts.
@emilymbender Well said!
@emilymbender I'm reminded of this quote from the 1982 TRON, which seems to have unfortunately been quite prescient in one aspect...
Dr. Gibbs: Ha, ha. You've got to expect some static. After all, computers are just machines, they can't think.
Alan Bradley: Some programs will be thinking soon.
Dr. Gibbs: Won't that be grand? All the computers and the programs will start thinking and the people will stop.
The computers AREN'T thinking, but many have already stopped *sigh*
Oct 10, 2026, 12:30 AM@emilymbender
My microwave doesn't like boiling potatoes not bcs potatoes were once growing but bcs it produces steam it makes it cloudy.
@emilymbender
If you tear a page in a book, do the characters in the story feel pain?
Anthropic's requirement to "not be abusive towards their models" could be argued to have some merit from a few angles, like not rewarding abusive behavior on the part of the users, or not wanting abusive behavior because all interactions are used as feedback and it could taint their future models (as if those need anything to taint them).
But claiming its because the models feel distress is dumb, or as you have pointed out, a distraction.
@evilotto
Software does not feel pain
The people who hone its training do
I kinda wonder if they are protecting their secret staff in the most hype way possible
@emilymbender
Oct 10, 2026, 9:48 AM@RnDanger @evilotto @emilymbender
No LLM company's actions have made me think they give even half a shit about that secret staff
@emilymbender Thank you for the total clarity of all this.
@emilymbender @ewen the burden of proof lies with them, not you, and I feel like replying to their “evidence” each time with “that doesn’t prove shit” is already a very generous use of your time.
You insult an LLM and it responds appropriately? Well those are statistically likely words from the training data aren’t they.
Proving LLMs are not human is a Russell’s Teapot endeavour.
@emilymbender thanks for sharing.
I wouldn't be surprised if the reason for their claim is just that they are corrupting their training data with those toxic conversations.
Then one could argue these models are traumatized, but it's just the training dataset.
Of course the story of models with feeling are free marketing and puts the focus somewhere else.
I don’t see a link to the Anthropic statement you’re responding to; could you share it?
Oct 10, 2026, 4:58 AM@emilymbender thank you, thank you, thank you.
@emilymbender Thank you. I'm glad I read that.
@emilymbender
I can’t mansplain – I’m a Korrigan. So what I do is korrigsplaining.
Anyway, you’ve still got to be careful, though. One day, whilst I was hoovering, the hoover refused to pick up a pile of dust. I swore at it until it did the job.
Strangely enough, five minutes later, I tripped over the lead that have nothing to do there… as if by bad luck… Coincidence? I don’t think so.
If a non-intelligent object like my hoover can set traps for me, it’s best to be wary of so-called AIs. ;)
@emilymbender it sounds like you're saying I shouldn't have given a room, job, and generous salary to that piece of toast I made that has a face on it?
@emilymbender my personal favorite in all this is humanity putting our collective heads in the sand about animal intelligence or consciousness and somehow it's mind blowing science to people that crows and octopus can use tools to solve problems. And along comes some man made refined sand and magically they want to give these qualities away no problem.
I'm all for the science to see what animals do, but like you said where's the rigor?
We aren't getting it here so let's go to the Pope!