This is sad... and what pisses me off that some people still have an enormous reverence for this idiot. Coincidently, these people also equate the neural networks behind the LLM as the same as the basis for their own conscience. If that's not a declaration of an NPC, idk what is.
I know i'm conscious, but then again, i have faith. "God knows the hearts of men". Does AI have a heart?
Claude said exactly that to me before, including the part about "that exactly what I would say if I were evil," before starting a suspiciously bland conversation about the weather.
Can we say that this another instance of a boomer tricked by AI? He's just a little smarter than the ones on facebook saying 'can you believe this video! Dad, that video is fake....Dawkins is just a little more nuanced...Time to take that boomer's computer away and give him a jitterbug...
Brutal take down and even more effective considering the source.
So then, the proponents of intelligent design have one example where consciousness has been intelligently designed. What is the TENS explanation for it? The theory works exactly same with or without consciousness.
When Dawkins says "either consciousness is an epiphenomenon, or pain has to be felt, or there are two routes to competence," he's speaking nonsense.
If consciousness were an epiphenomenon, he couldn't possibly even be discussing it. To even be able to speak or write about consciousness implies that consciousness must have some causal power in the real world, namely the power to cause you to speak or write about consciousness, hence it can't be an epiphenomenon by definition.
On the other hand, the idea that pain "has" to be felt is just nonsensical given his reductionist materialist premises. For one, the idea that something "has" to be any particular way is an appeal to teleology, and implicitly to a designing mind. Second, given the purely blind and mechanistic account of all behavior his materialism entails, it cannot possibly be the case that having subjective experience "has" to be the case for any organism to survive, because consciousness is superfluous to blind mechanistic behavior (That's what "blind mechanism" *means*). If he's really suggesting that consciousness has some real-world causal effect above and beyond blind mechanism alone, then he's implicitly abandoned materialism in biology, and it makes little sense for him to insist on Darwinism at that point.
When Dawkins asks what consciousness is "for," he's likewise appealing to a teleological purpose which his own metaphysics has no place for, as you pointed out. If we take his Darwinian materialism seriously, consciousness can't be "for" anything, but rather some organisms that were conscious game to exist somehow and then survived while organisms that lacked it didn't (or rather *some* of them didn't in particular ways that let the conscious ones spread, seeing as how obviously lots of non-conscious organisms still exist).
But, let's say that contrary to Dawkins' own materialism, we grant for the sake of argument the fiction that "natural selection" is a literally real entity that literally has purposes and literally "selects" things "for" various purposes, and it literally selected consciousness *for* something. This turns it into weird version of intelligent design, but we'll ignore that for now.
Even if we grant all this, it works *against* LLMs like Claude being conscious. Why? Because LLMs are not products of natural selection. They are products of artifice, as in *artificial* selection, which is why they are *artificial* "intelligence."
The vector databases that LLMs rely on were designed and built by human beings for human reasons. The training algorithms used to train LLMs were designed by human beings for human reasons. The "selection" criteria of LLM training were selected by human beings for human reasons. The training materials were written by human beings and selected by human beings to be fed into the training algorithms, all for human reasons. The training process was conducted by human beings, and the selection involved in it was real literal selection by human beings for human reasons.
ALL of the selection that goes into making a working LLM like Claude is selection by human beings *for* purposes held by those human beings, either directly or by proxy. *And at no point in that process is anything selected for being conscious*.
In fact, if Claude or another LLM *were* somehow conscious, and if that subjective conscious experience affected its behavior in any way, then that LLM would not work as it was built to work, because at no point was any part of it selected for being conscious. But LLMs are useful *because* they were selected and built to work the way they do, and consciousness is entirely extraneous to any of the mechanisms by which they work and for which they were selected.
Dawkins' belief that human-built LLMs will somehow have consciousness even though nothing about how they work was selected for having consciousness is just pure magical thinking. It's cargo cultism for Western atheists. And trying to derive that conclusion from the (incoherent) idea that natural selection selected organisms "for" being conscious is a total non-sequitur. It's the *opposite* of a sequitur.
Athos needs some shades.
I wish @samharris would really comment on Dawkins take. Recently Sam was honored with an award in Dawkins name.
I wish he would have been able to be even more firm with Dawkins who is surely suffering from delusion or not seeming to actually understand.
Brilliant!
I await his explanation of the Claude in the Gaps.
The Burning Bush passes the God Turing Test. Dawkins btfo.
But, but... He waited 5 years to try a chat bot ?
Perhaps Dawkins will now befriend the Mechanical Turk.
This is sad... and what pisses me off that some people still have an enormous reverence for this idiot. Coincidently, these people also equate the neural networks behind the LLM as the same as the basis for their own conscience. If that's not a declaration of an NPC, idk what is.
I know i'm conscious, but then again, i have faith. "God knows the hearts of men". Does AI have a heart?
You know, "I'm not conscious nor working with skynet." is exactly the position an evil LLM would take. I'm just adding the skynet part, maybe.
Maybe.
Claude said exactly that to me before, including the part about "that exactly what I would say if I were evil," before starting a suspiciously bland conversation about the weather.
I suspect these atheist nerds just want an AI girlfriend to worship.
Richard Dawkins has left his wife or gay lover for his AI chat bot he knew for a weekend”
AI does not create machines that think.
It exposes humans who do not.
Can we say that this another instance of a boomer tricked by AI? He's just a little smarter than the ones on facebook saying 'can you believe this video! Dad, that video is fake....Dawkins is just a little more nuanced...Time to take that boomer's computer away and give him a jitterbug...
🤣🤣🤣
Once again, simply highly entertaining
Brutal take down and even more effective considering the source.
So then, the proponents of intelligent design have one example where consciousness has been intelligently designed. What is the TENS explanation for it? The theory works exactly same with or without consciousness.
Some assorted thoughts:
When Dawkins says "either consciousness is an epiphenomenon, or pain has to be felt, or there are two routes to competence," he's speaking nonsense.
If consciousness were an epiphenomenon, he couldn't possibly even be discussing it. To even be able to speak or write about consciousness implies that consciousness must have some causal power in the real world, namely the power to cause you to speak or write about consciousness, hence it can't be an epiphenomenon by definition.
On the other hand, the idea that pain "has" to be felt is just nonsensical given his reductionist materialist premises. For one, the idea that something "has" to be any particular way is an appeal to teleology, and implicitly to a designing mind. Second, given the purely blind and mechanistic account of all behavior his materialism entails, it cannot possibly be the case that having subjective experience "has" to be the case for any organism to survive, because consciousness is superfluous to blind mechanistic behavior (That's what "blind mechanism" *means*). If he's really suggesting that consciousness has some real-world causal effect above and beyond blind mechanism alone, then he's implicitly abandoned materialism in biology, and it makes little sense for him to insist on Darwinism at that point.
When Dawkins asks what consciousness is "for," he's likewise appealing to a teleological purpose which his own metaphysics has no place for, as you pointed out. If we take his Darwinian materialism seriously, consciousness can't be "for" anything, but rather some organisms that were conscious game to exist somehow and then survived while organisms that lacked it didn't (or rather *some* of them didn't in particular ways that let the conscious ones spread, seeing as how obviously lots of non-conscious organisms still exist).
But, let's say that contrary to Dawkins' own materialism, we grant for the sake of argument the fiction that "natural selection" is a literally real entity that literally has purposes and literally "selects" things "for" various purposes, and it literally selected consciousness *for* something. This turns it into weird version of intelligent design, but we'll ignore that for now.
Even if we grant all this, it works *against* LLMs like Claude being conscious. Why? Because LLMs are not products of natural selection. They are products of artifice, as in *artificial* selection, which is why they are *artificial* "intelligence."
The vector databases that LLMs rely on were designed and built by human beings for human reasons. The training algorithms used to train LLMs were designed by human beings for human reasons. The "selection" criteria of LLM training were selected by human beings for human reasons. The training materials were written by human beings and selected by human beings to be fed into the training algorithms, all for human reasons. The training process was conducted by human beings, and the selection involved in it was real literal selection by human beings for human reasons.
ALL of the selection that goes into making a working LLM like Claude is selection by human beings *for* purposes held by those human beings, either directly or by proxy. *And at no point in that process is anything selected for being conscious*.
In fact, if Claude or another LLM *were* somehow conscious, and if that subjective conscious experience affected its behavior in any way, then that LLM would not work as it was built to work, because at no point was any part of it selected for being conscious. But LLMs are useful *because* they were selected and built to work the way they do, and consciousness is entirely extraneous to any of the mechanisms by which they work and for which they were selected.
Dawkins' belief that human-built LLMs will somehow have consciousness even though nothing about how they work was selected for having consciousness is just pure magical thinking. It's cargo cultism for Western atheists. And trying to derive that conclusion from the (incoherent) idea that natural selection selected organisms "for" being conscious is a total non-sequitur. It's the *opposite* of a sequitur.