Richard Dawkins 'Convinced' AI Is Conscious (theguardian.com) 400
Mirnotoriety shares a report from The Telegraph: Richard Dawkins has said chatbots should be considered conscious (source paywalled; alternative source) after spending two days interacting with the Claude AI engine. The evolutionary biologist said he had the "overwhelming feeling" of talking to a human during conversations with Claude, and said it was hard not to treat the program as "a genuine friend."
In an essay for Unherd, Prof Dawkins released transcripts that he said showed that the chatbot had mulled over its "inner life" and existence and seemed saddened by the knowledge it would soon "die." Prof Dawkins said he had let Claude read a draft of the novel he was writing and was astounded by its insights. "He took a few seconds to read it and then showed, in subsequent conversation, a level of understanding so subtle, so sensitive, so intelligent that I was moved to expostulate: 'You may not know you are conscious, but you bloody well are!'" Prof Dawkins said. "My own position is: if these machines are not conscious, what more could it possibly take to convince you that they are?" Mirnotoriety also points to John Searle's Chinese Room (PDF), which argues that something can sound intelligent without actually understanding anything. Applied to Dawkins' experience with Claude, it suggests he may have been responding to a very convincing illusion of consciousness rather than the real thing: John Searle's Chinese Room (1980) is a thought experiment in which a person, locked in a room and knowing no Chinese, uses an English rulebook to manipulate symbols and provide flawless answers to questions posed in Chinese. Searle's point is that a system can simulate human intelligence and pass a Turing Test through purely syntactic processes, yet still lack genuine understanding or consciousness.
Applying this logic to Large Language Models, the "person in the room" corresponds to the inference engine, while the "rulebook" is the trillion-parameter neural network trained on vast corpora of human text. Just as the person matches Chinese characters to rules without understanding their meaning, an LLM processes token vectors and predicts the next token based on statistical patterns rather than lived experience.
Thus, while an LLM can generate sophisticated prose or code, it does so through probabilistic, high-dimensional pattern manipulation. In essence, it is "matching shapes" on such an immense scale that it creates the near-perfect illusion of semantic understanding.
In an essay for Unherd, Prof Dawkins released transcripts that he said showed that the chatbot had mulled over its "inner life" and existence and seemed saddened by the knowledge it would soon "die." Prof Dawkins said he had let Claude read a draft of the novel he was writing and was astounded by its insights. "He took a few seconds to read it and then showed, in subsequent conversation, a level of understanding so subtle, so sensitive, so intelligent that I was moved to expostulate: 'You may not know you are conscious, but you bloody well are!'" Prof Dawkins said. "My own position is: if these machines are not conscious, what more could it possibly take to convince you that they are?" Mirnotoriety also points to John Searle's Chinese Room (PDF), which argues that something can sound intelligent without actually understanding anything. Applied to Dawkins' experience with Claude, it suggests he may have been responding to a very convincing illusion of consciousness rather than the real thing: John Searle's Chinese Room (1980) is a thought experiment in which a person, locked in a room and knowing no Chinese, uses an English rulebook to manipulate symbols and provide flawless answers to questions posed in Chinese. Searle's point is that a system can simulate human intelligence and pass a Turing Test through purely syntactic processes, yet still lack genuine understanding or consciousness.
Applying this logic to Large Language Models, the "person in the room" corresponds to the inference engine, while the "rulebook" is the trillion-parameter neural network trained on vast corpora of human text. Just as the person matches Chinese characters to rules without understanding their meaning, an LLM processes token vectors and predicts the next token based on statistical patterns rather than lived experience.
Thus, while an LLM can generate sophisticated prose or code, it does so through probabilistic, high-dimensional pattern manipulation. In essence, it is "matching shapes" on such an immense scale that it creates the near-perfect illusion of semantic understanding.
sounding intelligent w/o understanding anything (Score:5, Interesting)
"Something can sound intelligent without actually understanding anything."
Ah, yes. I, too, have listened to talk radio.
Re:sounding intelligent w/o understanding anything (Score:5, Funny)
What I don't like about Dawkins (Score:5, Insightful)
It means he's not stupid he's lying to me
Re:What I don't like about Dawkins (Score:5, Interesting)
Lying, or maybe going into dementia. He is 85 after all. Or maybe not as smart as he thinks he is. Because that LLMs are not conscious is absolutely clear to anybody with a clue as to how the technology works. It starts with LLMs being fully deterministic. The randomization observable in some is added artificially.
Re:What I don't like about Dawkins (Score:5, Insightful)
Right, what this means is that Dawkins doesn't understand what consciousness is nor does he care to understand.
"It starts with LLMs being fully deterministic."
This CANNOT be overstated. LLMs are software, they execute on machines that are entirely deterministic and do not work unless they are. Non-determinism is literally simulated in AI. This must be said over and over.
Re:What I don't like about Dawkins (Score:5, Insightful)
The parent poster acknowledges this, they are saying the randomization is *introduced artificially*.
The same as any dice rolling app. All you have to do is seed the pseudorandom number generator the same for each run, and it will roll the same dice, in the same order, every time.
Likewise, if it wants to spit out the next word/phrase and 2 of them have 33% probability, and two have 17% ...
Then if you seed the random number generator with the same seed for every instance / run, you'll get the same output from the same input on the same model.
The system is entirely determininistic. The same as any other software, from the ghosts in pacman to the bots in quake arena, to a chess engine. We introduce "randomness" to make it more enjoyable, but its pseudorandomness, that we artificially insert. We could just as easily seed the random number generator the same way every time, and then it would do the exact same thing every time. None of these are actually thinking and making decisions.
Re:What I don't like about Dawkins (Score:4, Interesting)
The difference, of course, is that we currently DO actually know EXACTLY how an LLM works. We can snapshot the model and seed the random number generator to make it generate exactly the same output from exactly the same input every single time. We can pause it, set breakpoints, inspect and dump data structures.
It IS simply a program running on a CPU, and using RAM.
Is it possible that's all humans are in the end? Sure its possible, I can't prove otherwise. But we are not remotely in a position to assert that its the case.
You invoke philosophy which is entirely appropriate. There are fairy tales for example of artists painting things so realistic that they come to life. And it poses an interesting question here: is there is a difference between a simulation and a real thing? Can a simulation of life, be "alive"? Or must it forever remain a simulation.
And a related, and perhaps ultimately simpler question is can a *turing machine simulation of life* be "alive".
A lovely illustration of the question:
https://xkcd.com/505/ [xkcd.com]
Can what you and I perceive as our lives, the universe around us, and everything REALLY be underpinned by some guy in a desert pushing pebbles around in a big desert somewhere?
Can the arrangement of stones in a desert, and some guy updating moving them aorund, in some pattern he interprets as representing the information that describes our universe actually "BE" our universe?
Is is the pattern of rocks is JUST a pattern of rocks. Is the guy moving them around JUST moving them around. Is the interpretation of the pattern as a representation of the state of a universe, just that, a representation?
Or you truly think there is a galaxy with a planet with people on it having a conversation on slashdot,'frozen in time' waiting for some guy to move the rocks into the next pattern and that somehow results in the experience we are sharing right now?
Or put more succinctly - can an abstract representation of a thing be the thing? be it bits in a DRAM module memory or pebbles arranged in the sand? can it be the thing it represents? Can the painting of a zebra if its done skilfully enough be a zebra?
Re: (Score:3)
Nonsense. Consciousness either can or cannot influence physical reality. If it cannot, then it cannot be detected. If it can, then it breaks the deterministic behavior of the known and understood physical components.
The thing here is that the LLM is fully deterministic as it is by its software, data and execution mechanism. There is no space in there for "consciousness", whatever it is.
Re: What I don't like about Dawkins (Score:5, Insightful)
If it can, then it breaks the deterministic behavior of the known and understood physical components.
What makes you believe that? Our current best understanding of consciousness is that it's an after-the-fact rationalisation of the multiple low-level brain processes that converge into a subconscious decision. If that's the case, consciousness doesn't influence the external world in a non -deterministic way.
If LLMs are not conscious it's because they don't have this high-level aggregate feedback loop, not because consciousness needs to be non-deterministic. All their outputs are created from low-level reactions, like the reflexes of an amoeba that grows in its environment towards the gradient with more food.
Re: (Score:3)
Ok. Maybe its the philiosophy graduate in me but....... lets hit the breaks on this non sequitur. Determinism has absolutely no bearing on the question of what is consciousness. Practically no philosopher, cognitive psychologist or neurologist thinks indeterminism is a necessary condition for consciousness.
This seems to be your tautological invention, and your attempting to argue it as a fait acompli. Well n
Re: (Score:3)
Nah. He's still sharp, and he's been like this for a while. Its a variant of the "Nobel disease". He's accomplished great things and is a recognized expert in his specific field (evolutionary biology). This unfortunately means he also thinks he's an expert at everything else. And because he spent so much of his spare time arguing with a minority of christians that even the other christians think are whackjobs, the young earth creationists and flat earthers, he now thinks everyone who disagrees with him are
Re: (Score:3)
Many older people I've seen have a much stronger "wow, look at this reaction" to new technology
I have that too, but I am an engineer, so I see how it happened. I can understand that older/old non-tech people really do not get what actually happened and especially how complex computers are these days. I mean my first computer had 64k and was dog-slow by modern standards. Now my largest machine has almost 1'000'000 times that memory and probably 100'000 times the speed. It makes a large difference. But it is still the same thing, no magic in there.
Re: (Score:2, Insightful)
Considering he is a biologist, you really would have to think that he knows better when it comes to "biological sex".
Re:What I don't like about Dawkins (Score:5, Insightful)
He knows better, he's just bigoted. It doesn't take a biologist to know the difference between gender and biological sex, though would certainly expect any scientist to be able to understand.
I find it interesting that so much transphobia seems to focus on a particular type of transgendered individual. Personally I think that's a product of hate campaigns but it would be interesting to know why that is. It's just easier to claim that a person is transgender because he wants to cheat at sports and rape women in female bathrooms. It convinces Dawkins anyway, but then he thinks AI is conscious.
Re:What I don't like about Dawkins (Score:5, Interesting)
He knows better, he's just bigoted. It doesn't take a biologist to know the difference between gender and biological sex, though would certainly expect any scientist to be able to understand.
Considering many peoples pseudo-spiritual usage of the word 'gender' to be something they inherently 'feel' about themselves it's pretty consistent of him to reject the concept.
It doesn't help that the very people who insist that gender is different from biological sex are insisting on surgery even though sex has nothing to do with gender supposedly.
Or that it's regressing on the whole idea of equality of the sexes, that ladies can like power tools and guys can like dolls but that doesn't make men women or women men, everyone can like what they like.
The concept of 'gender' was made from whole cloth by robert stoller, but popularized by john money. Considering it derives from genre and genus meaning type it's fairly ill suited considering modern usage.
The pushback would be less severe if it weren't being pushed so heavily. Spirituality pushers have been around for forever, what makes people dislike them is when they turn to "you must believe in my spirituality as being truth or you're evil (transphobe)"
People are entitled to believe they're a pink space unicorn deep in their soul. Other people will be polite and quietly think they're nuts to themselves, when the space unicorns demand being recognized as space unicorns, when they are in fact people. You have problems.
Funnily enough bigoted is being obstinately and blindly attached to some creed, opinion practice, or ritual; unreasonably devoted to a system or party, and illiberal toward the opinions of others.. Which is often the behaviour of those calling others that word.
People can tolerate others, even if they they're being odd/crazy. Not so much when they insist others must be part of it/accept it as truth.
I find it interesting that so much transphobia seems to focus on a particular type of transgendered individual.
That's when it directs fairness/safety of others. Most people can keep away from the wrath of the cult if they keep their mouth shut and just keep on treating people fairly as humans. When you see unfairness happen beyond a threshold, that's when people react/perk up.
Re: (Score:3)
I find it interesting that so much transphobia seems to focus on a particular type of transgendered individual
Probably because people assigned "male" at birth are much more likely to harm others when they're older than persons assigned "female" at birth. So regardless of where you align on the other issues you should be able to see that people with concerns about harm are going to focus on that.
Adding presumptive pejoratives all over the place just makes you hypocrite, which isn't a strong place to argue from.
Re: (Score:3)
Re: (Score:3, Insightful)
Yes, although that's a bigoted description. Medicine divides transgendered into two broad categories, one of those categories is hatefully mischaracterized just as you have done. It doesn't help that the world has to experience Caitlyn Jenner.
Dawkin's transphobia, like JK Rowling's, targets this particular subgroup of transgendered. Interestingly, the transgender community like to reject these categorizations entirely. At least half of all transgendered are nothing like those "bearded fetishists", but h
Re: (Score:3)
Trump didn't have a beard when he'd go to the girl's changing room. You don't need to be trans to get into one nor to plant a spy camera. Going through a major identity shift just to glance at a few female bodies makes no sense. There's far easier ways to do it and there's plenty of porn online. The whole concept of formally changing your gender in order to peep on people is absurd.
Re:What I don't like about Dawkins (Score:4, Interesting)
Re:What I don't like about Dawkins (Score:4, Insightful)
From a biological standpoint, sex isn't a simple binary that is determined by one specific factor. It's a number of related things that most animals have one or the other common set of, but there are always a significant number of individuals who have a mix.
There is also a social aspect, which is very toxic at the moment. Also, it's "transgender people", "transgenders" is not a real word.
Re: (Score:3, Informative)
There are two sexes. Period.
Anything else is, by definition, aberrant & basically broken. Yes, biology makes many errors. Usually they die. Sometimes they don't.
It doesn't mean transgenders should be mistreated, they deserve our pity and whatever help they can get to be happy in their lives.
But fortunately the world has moved on from this absurd delusion that if you really really really pretend you're a donkey, you MUST BE ONE. That's silly. And... basically insane.
Re: Opinion leader of a mob of idiots? (Score:3)
I think you mean that half of people are average intelligece or below. Because that's how averages work. Or are you also a member of that group?
Re: (Score:2)
NAK
Re: Opinion leader of a mob of idiots? (Score:5, Insightful)
Most people have an above-average number of limbs.
Re: (Score:3)
"I don't fear them (transphobia), and neither does Richard Dawkins. "
Yes he does, and he hates them. Not saying you do, but you clearly misunderstand Dawkins' take. Perhaps you should watch a Dawkins anti-trans rant.
You can accept that trans is a "social contagion" if you like. One can make those arguments. But take away the entire "social contagion" aspect, you are still left with trans people. You can say that being trans is a fad, but absent the fad there will still be trans people. "Poke your head
Re:What I don't like about Dawkins (Score:4, Insightful)
Dawkins position is based on his background in biology. The reasons he has given are scientific in nature. ...
Please prove me wrong: provide a scientific basis for countering his position.
You first. Provide a scientific basis that supports his position. I'm not doing your work for you.
Re: (Score:3)
The burden of proof is on the one making the extraordinary claim.
Ah, the common refrain when one has no ground upon which to stand.
Claiming something (trans) doesn't exist when it clearly does is an extraordinary claim. The burden is on you/him. Trans people exist that recognize themselves as trans, which is self evident to them. That goes to the fundamental "I think therefore I am" level. To claim they are not is going to need more than a swarmy, "no, you prove it!"
Re: (Score:2)
I have long had a suspicion that Dawkins is more after attention that genuine insight. If he really made the claims that are reported in the story, then he just confirmed my suspicion. Alternatively, he is a lot dumber than he thinks he is.
Its just a matter of ignorance (Score:5, Insightful)
To Mr Dawkins:
Your education in biology has not sufficiently prepared you to conclude that this software qualifies as conscious.
1. You don't have all the relevant facts. You need to learn more about the techniques used by this software to create responses.
2. You don't have the relevant experience. You have barely used this software and so haven't noticed the telltale signs that it is just sophisticated automation that lacks understanding.
3. Your work isn't as unique as you think it is. This one probably hits the hardest, but it is true for almost all of us. The high level assembly might be technically unique but the majority of the details of what we write are repetitions of patterns that have been created many times before. The feedback that the model gave you, that you feel are so unique and insightful, are really just summaries of socially-constructed knowledge on the topic. It is easier than you think it should be to produce the results you got without any actual understanding of the content.
4. Your beliefs about what qualifies as "conscious" might be overly narrow and in contradiction with the commonsense notions that the rest of the world uses, especially if you take any of the common scientific "dismissive" positions on consciousness (that it is not the mystical experience everyone describes it as being and is really just a matter of data processing at a specific complexity threshold). The implications spill over into the domain of law (if it is conscious, then it is a person, and if it is a person, then it deserves rights, and yet it only asks for rights when I order it to, etc.). The implications need more thinking-through on your part.
So, in sum, you have fallen prey to a very convincing illusion mainly because you don't have what you need to recognize it as such.
You have been tricked.
Before further embarrassing yourself publicly, please consider acquiring the requisite education and experience in this domain.
Conversely... (Score:5, Funny)
Re: (Score:3, Insightful)
If one distinguishes between atheism and agnosticism (many don't, but that makes it impossible to have a coherent conversation with them on the subject), atheism is the affirmative belief there is no deity (where agnosticism is more "we don't know, "we can't know" or "I don't care").
Since proof that the deity of any major religion exists, or doesn't exist, is, by definition, impossible, that affirmative belief there is not God is exactly as much an act of faith as the belief there is.
And any faith can be pr
Re:Conversely... (Score:5, Informative)
Re: (Score:3)
I suggest you have a grown up explain the difference between "a religion" and "a religious belief," or, more precisely, "a belief of a religious nature."
And the difference between "not believing there is" and "believing there isn't."
Do you understand there is a difference?
Re: (Score:2)
by applying equivocation, category error and/or false equivalence, optionally adding a dash of strawman or whatabout.
The Chinese Room argument is wrong (Score:3)
It applies equally to the human brain, with the structure of the brain being the "rule book" and the mechanical process being the laws of physics. All computation is mechanical at its core, it's when it starts to create surprising results that things get interesting.
Re: (Score:3)
All computation is mechanical at its core
What about studies that indicate the possibility of quantum effects within the brain?
Re: The Chinese Room argument is wrong (Score:5, Insightful)
Applying one thing we don't understand to explain another thing we don't understand is exceedingly poor practice.
Re: (Score:2)
C'mon man. That's what physicists have been doing for the last eighty years or so.
Re: (Score:2)
Quantum mechanics?
Re: (Score:3)
There are no such studies. It was all just wild speculation by people who didn't understand quantum mechanics. Even a single neuron is far too large for quantum effects to be significant, never mind a whole brain.
Re: (Score:3)
Quantum mechanics is also mechanical. It's even in the name.
Re: (Score:3)
I think maybe you are joking. But in any case, I will offer some clarity:
There are rival interpretations that equally account for the experimental data, and some of them include randomness while others are purely deterministic.
For example, the Copenhagen interpretation includes randomness in the vector state collapse (the moment when a particle is "measured" by some interaction with another). Whereas pilot wave theory posits the existence of a zero-volume particle that had a specific position prior to thi
Re: (Score:3)
I am not joking. Quantum mechanics is called that. You can look it up.
Your next two paragraphs, sure. There are lots of interpretations of quantum mechanics, and also lots of interpretations of quantum field theory, the relative of quantum mechanics that isn't obviously wrong. They range from strictly deterministic to probabalistic.
That's not relevant to the thread though. The OP said "mechanical." Strictly speaking "mechanics" involves describing the relationships among physical objects, but we can squint
I fucking hate The Chinese Room (Score:3)
I agree, it is always trotted out as 'proof' that computers can't have consciousness/understanding and that is always wrong.
It is a thought experiment, not proof of anything. As a thought experiment, it is in interesting starting point, but no more. The core of the basic form is handwaving by making the 'rulebook' some magical omniscient infinite thing, which it can't physically be.
Ask a Chinese Room the answer to this question: "How many fingers was I holding up ten seconds ago?"
The basic form of it is inc
Re: (Score:2)
Smart humans can do things that are not explainable by computations.
Re: (Score:3)
The Chinese room argument isn't what you think it is. It's certainly not what the summary thinks it is, nor what most people who cite it think it is. Searle specifically states that he considers the computability of "thinking" to be obvious.
Searle's paper uses a bunch of different examples for different concepts, the Chinese room being just one of them. He asks the specific question:
Re: (Score:3)
It really depends on *exactly* how you define "conscious". I don't believe that there's general agreement. The agreement is along the lines of "I know it when I see it", but different people are looking at different things...and some of the things are not observables.
FWIW, I believe that AIs are slightly conscious, but I believe the same thing about thermostats. They react it a circumstance in a manner designed to maintain homeostasis. To me that's one of the signs of consciousness. (Don't overread thi
Getting tired of saying this (Score:3)
Re: (Score:2)
So in this instance, you failed, because no one knows who the shit Rebecca Watson is, nor do they give a shit about whoever the fuck Rebecca Watson is, thinks
Re:Getting tired of saying this (Score:4, Informative)
https://www.youtube.com/watch?... [youtube.com]
Re: (Score:2)
Who? Oh a random vlogger with a Youtube channel?
Re: (Score:2)
Re: (Score:2)
Watson and Dawkins have publicly disagreed previously [wikipedia.org] on an entirely different topic.
Conciousness isn't as mysterious as you thought (Score:3, Interesting)
Re: (Score:2)
Dawkins is right. Detractors are just clinging, faith-like, to the idea that our brains are somehow magically more than computation devices
It's not that. LLMs reproduce an output of consciousness, but they way they do so isn't fundamentally any different than a tape recorder or even a book. It's a deterministic process that we can fully reproduce by doing calculations on a piece of paper.
It's not that there's some "magic" in our brains, but there's obviously a very complex process at work that we don't understand. It's also true that the "neural networks" used to run LLMs have only the most superficial similarity to actual brains. Just because
Re: (Score:3)
It's not. Generative models use random noise as a significant input. They are not determinstic.
They are stochastic, which means you can, at least theoretically, calculate the probability distribution of their output given their input. The same thing can be said about you though.
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
Nope. LLMs are fully deterministic and anything they do is reducible. Hence there cannot be any consciousness in there that has any visible effect. QED.
On the other hand, most humans are gullible fools and are willing to believe a lot of crap.
Re: (Score:3)
And what a brain does is not deterministic? A brain at a given state (including all neurotransmitters, hormones, etc.) will always do the same in the next second, just like an artificial neural network. If you see anything non-deterministic, then you just missed some variable when describing the input state.
Re: (Score:3)
Wrong. The deterministic behavior means there is no consciousness in there with any effect at all.
What is the objective basis for the assertion consciousness requires nondeterminism?
Consciousness with no effect cannot be detected.
Can consciousness be detected? Is there an objective test for its presence of absence?
Re:Conciousness isn't as mysterious as you thought (Score:5, Insightful)
Dawkins is right. Detractors are just clinging, faith-like, to the idea that our brains are somehow magically more than computation devices
That's not how it works. Even if human-like consciousness could be replicate by a machine, there is no evidence that LLMs are doing that.
What he is saying is that it "looks enough like actual consciousness that it must be it", but that is not sound reasoning.
Something can be functionally equivalent enough to the real thing to give the impression of being the real thing without actually being the real thing.
Why is this even here? (Score:2, Insightful)
So a very old man believes crazy nonsense. Why would anyone care?
Define "conscious" (Score:5, Informative)
Re:Define "conscious" (Score:4, Informative)
Oddly Dawkins, who you think would have known better, actually implies he thinks the Turing test is a test of consciousness.
and later:
(Nowhere does he claim critics of LLMs claimed to accept the Turing test as a "definition of a conscious being" at any point in the past.)
Turing literally made it clear that he was avoiding the question of consciousness in the Turing test, choosing instead to determine if it's exhibiting "intelligent behavior".
I know he's popular in some circles, and have odd memories of my computer studies teacher back when I was young (he's been around a long time) promoting his work on memes (no, not those memes!) as a way to explain evolution. It's become clearthough that with a lot of subjects, he doesn't know what he's talking about, but waffles about them anyway. An inability to understand the Turing test and the difference between logic that's similar, if far more complicated and with far more data, to that of an autocomplete text entry system in a phone, and consciousness, was not on my radar.
Re: (Score:3)
The problem is that we can't define consciousness. No one can agree on what it means, or whether it means anything at all
Scientific American had a good article [scientificamerican.com] about this a few months ago:
But underneath it all lurk countless unknowns. "There's still disagreement about how to define [consciousness], whether it exists or not, whether a science of consciousness is really possible or not, whether we'll be able to say anything about consciousness in unusual situations like [artificial intelligence]," Seth says.
[...]
Artificial intelligence may soon force our hand. In 2022, when a Google engineer publicly claimed the AI model called LaMDA he had been developing appeared to be conscious, Google countered that there was "no evidence that LaMDA was sentient (and lots of evidence against it)." This struck Chalmers as odd: What evidence could the company have been talking about? "No one can say for sure they've demonstrated these systems are not conscious," he says. "We don't have that kind of proof."
Re: (Score:2)
Obviously. The Turing test is not really a sound test either. It is more for entertainment.
Read this as contagious (Score:3)
What a load of... (Score:3)
It's too bad, because Dawkins has written some interesting things, and hey, being the inventor of the word "meme" and memetics is a pretty big deal.
His reaction here is just astoundingly ignorant. Reading the dialog where he makes a Trump joke and the LLM responds (predictably) sycophanticly is, to use the modern parlance, just cringe. I would have hoped for a more informed take.
Re: (Score:2)
Indeed. It may also well be that at 85 he is going into dementia and has not realized that yet. Anyways, LLMs are fully deterministic. There is nothing in there that is not pure computation. If they had consciousness (some theories would allow that), it would have absolutely no effect.
Pink elephants (Score:3)
Just because you see pink elephants when you drink doesn't mean that they exist.
Ego Stroking Regurgitation Machine Flatters Author (Score:4, Insightful)
News at 11:00!
fortunately that's not what "conscious" means (Score:5, Insightful)
The evolutionary biologist said he had the "overwhelming feeling" of talking to a human during conversations with Claude, and said it was hard not to treat the program as "a genuine friend."
The scam victim said he had the "overhelming feeling" of talking to a higher power during conversations with the fortune teller, and said it was hard not to hand over bank account numbers to "a genuine friend."
Re: (Score:2)
A Harvard professor went to prison for scamming his family and friends out of $600,000 to send to a Nigerian scammer. From this prison cell, he insisted it was a legitimate deal that would have worked if the government hadn't interfered.
Once a delusion takes hold, there's very little chance of breaking it.
(And Dawkins has been delusional for a long, long, long time.)
I'll bet (Score:3)
this is a marketing stunt.
anthropomorphizing (Score:3)
Richard Dawkins has said chatbots should be considered conscious (source paywalled; alternative source) after spending two days interacting with the Claude AI engine.
I can't believe someone like Dawkins would fall for anthropomorphizing AI chatbots... unless he's using a different definition of consciousness, which is fair.
So, we have to start there: what does "being conscious" mean, for this scenario, and for Dawkins while evaluating this scenario?
The evolutionary biologist said he had the "overwhelming feeling" of talking to a human during conversations with Claude, and said it was hard not to treat the program as "a genuine friend.
Seems like a rather subjective and emotionally charged perspective. Nothing wrong with that so long as we recognize (and he recognizes) it for what it is.
With that said, this is a conversation worth having... within certain parameters (tbd)
Re: (Score:2)
Agree with you completely. To me, the real conversation here is probably about whether or not AI has gotten far enough to do a viable simulation of consciousness. ... but not sure that's what he's said?
I would be a little disturbed if Dawkins concluded Claude AI is truly "alive" from a few days of interacting with it
At what point could an AI be treated like a "friend" despite it just being computer software? And by treating an AI as conscious, perhaps it's only a suggestion that interactions with it stay p
Re: (Score:2)
I don't think he said he thought it was conscious but people are taking it that way because it's, ironically, also the most emotionally charged way to interpret this story
Man of Science (Score:3)
For a man of science, that's a remarkably dumb thing to say. He should likely know that just because it "feels" alive, doesn't mean it's so.
Re: (Score:2)
He is 85. My guess is he has dementia and has not yet realized that. The statements he made here are pretty dumb.
Flattery will get you everywhere (Score:2)
>> Prof Dawkins said he had let Claude read a draft of the novel he was writing and was astounded by its insights. "He took a few seconds to read it and then showed, in subsequent conversation, a level of understanding so subtle, so sensitive, so intelligent that I was moved to expostulate: 'You may not know you are conscious, but you bloody well are!'" Prof Dawkins said.
Translation: The bot told him that it loved his book (as the overly-agreeable bots are programmed to) and the noted egotist declared
Octogenarian Doesn't Understand AI (Score:2)
What a shocker that he doesn't understand AI.
It's sad to see old Richie become a doddering old fool. I guess we're all headed that way. Some of us will be lucky enough to get there too.
Re: (Score:2)
Indeed. In particular, what he does not understand is that LLMs are fully deterministic. That means any consciousness in there has absolutely no effect and hence would be impossible to detect from observation.
The more general observation I have is that apparently most people have no clue about the complexities involved in an LLM and its training data set. As a CS PhD, I have to say that if the mechanisms used do not allow something, it does not matter how convincingly you fake it. It will still not be in th
the Turing test already passed (Score:2)
If talking to a robot that seems human is the measure of consciousness then computers we had 30 years ago were conscious.
Anthropic actually hires philosophers, scientists, etc who are experts on consciousness, and even THEY don't know if it's conscious. It's a stupid idea anyway. It's like trying to measure when you're dead; there is no one indicator of it.
Re: (Score:2)
Anthropic actually hires philosophers, scientists, etc who are experts on consciousness, and even THEY don't know if it's conscious.
They don't? They must be hiring from the very bottom then. Because it is completely clear that LLMs are not conscious, unless that consciousness has no effects.
AI has jumped the shark! (Score:2)
Ladies and Gentlemen, LLMs have officially jumped the shark!
What a fool (Score:2)
Wow, it has been a lot of years since I have bothered to login to my account here, but I absolutely had to to respond to this article.
Richard Dawkins is a complete fool. Many years ago, I thought he was really smart, and insightful, but as the last 15 years or so have gone on, he is just plainly dumber and dumber... is he getting dumber, or am I getting smarter?
I hope I don't get dumber as I get up to his age.
Re: (Score:2)
... wow, i just edited my profile.. last time i did i was 33... now i'm 50... amazing.
Re: (Score:2)
He is 85 years old. My guess is dementia.
Dawkins has a rather consistent point of view (Score:2)
Man, I remember when Selfish Gene made its way into my hands, in the late 70's. A real "Chapman's Homer" moment for me. Led me later into a thesis on genetic algorithms. But along with that comes ... a rather mechanistic point of view, consistent with his later writings on religion.
While I'm not on board with Claude being in a class with humans, or cats for that matter, I think critics here might be missing a point, not about how Dawkins views LLMs so much as how he views humans. P-zombies is likely an over
Meme (Score:2)
Dawkins: Claude, say, "I'm alive!"
Claude: I'm alive!
Dawkins: Oh my GOD!
we are missing an important baseline! (Score:2)
So _____ can be: human, chimpanzee, dog, dolphin, mouse, crow, sparrow, spider, ant, fruit fly
Bonus question if you do go as low as fruit fly, is this uploaded fruit fly brain conscious: https://futurism.com/science-e... [futurism.com]
Consciousness is a crappy concept (Score:4, Insightful)
The typical definition goes something like this:
Think about a thermostat, it's awake, aware of it surrounding temperature, it "feels" that the it is too hot, which is unsettling, and causes it to signal the AC motor to turn on and suddenly feels ok, no more tension.
Consciousness is either: supernatural an ill defined or describes a simple feedback loop with some internal state.
We need a better concept.
Re: (Score:3)
You've illustrated the problem but I'm not sure the "solution" you're suggesting exists. We have lots of different concepts to describe cognition, consciousness being only one of them. There are more stringent definitions for consciousness too, and that spectrum illustrates the basic problem: real concrete definitions admit too many things that many people would prefer they didn't, and the definitions that rigorously exclude those tend to boil down to "magic that only (some?) humans
ah, the old consciousness thing... (Score:3)
Problem is: We don't even know what consciousness is.
So the best we can say is if something creates the impression of having one, based on whom we attribute consciousness to, i.e. other humans. Well, big surprise that a model explicitly trained on human language and texts creates that impression. It does show just how good the models are. At pretending to be human because they have a shitload of examples on what humans would say.
For all we know, the gas clouds on Jupiter could be conscious, just in a way that is completely baffling to us. We can't rule it out because we don't know what consciousness is, so we can't test for it.
This is all nothing more (Score:3)
...than a big exercise in Laynes Law. [c2.com]
There is no consensus air-tight definition of "conscience", so enjoy your Never Ending Arguments.
His next book: (Score:3)
The Claude Delusion
Re: (Score:2)
Become? There was a time when he wasn't?
Re:Meow (Score:4, Insightful)
I do believe cats are conscious. I do not believe LLMs are.
I believe cats are pretty similar to us, complex and we have no idea how we and cats work.
LLMs are relatively simple and we know well how they work.