I Get Email
Posted in Blogosphere, Ethics, Web/Tech on | 5 minutes | 31 Comments →
A few days I go I received an email from bossmanham:
I took Luke’s blog off of my Google reader feed a while back because he was getting boring and predictable, but moseyed on over yesterday. I probably shouldn’t have been, but I was a little surprised that he’s focusing so much time on machine ethics. I guess that’s what you get when inventing your own ethics, but I was wondering what your thoughts were since you’ve followed his blog longer than I have?
Where to begin?
I suppose I should start from a writer’s perspective and give Luke credit for writing about what he feels is most important. Like I said, I admire the fact that he’s writing for himself. I’ve never been impressed with writers who value pleasing their audience over writing from the heart. To me, this shows that Luke is more interested in finding and following his own voice than maximizing his readership, and I say bravo to that.
That said, I agree that Luke’s gotten boring and predictable. Truth be told, after what I would describe as a stellar beginning at Common Sense Atheism, I thought Luke had gotten boring and predictable by the end of 2009, ever since he ponied up with Alonzo Fyfe. How can I not be at least a little let down when the blogger who started with such seemingly genuine respect for conservatively stated claims declares that his newfound beliefs in materialism and atheism are “settled issues?” Just a few short years ago, Luke would have told you they were “settled issues” in favor of God’s existence and dualism. Personally, I’m a fan of the tortoise, not the hare, and it seems to me that Luke has simply traded his cross for a scarlet A, swapped one faith for another. If there’s one statement D forever etched into my mind, it’s that everybody needs a gris-gris.
As a hobbyist programmer, the topic of machine ethics isn’t 100% devoid of interest for me, but I’m just not that interested in reading about it. I agree with Luke that the topic is of paramount importance to the survival of our species, but the whole thing is a bit like politics for me: a runaway train completely out of my control, a bull in a china shop thrashing rampantly amidst the fragility of organic life. Apathetic? Perhaps, but I tend to think I’m being realistic. Sure, I want these lunatics to stop, but what am I going to do? Get all Ted Kaczynski and send AI researchers care packages in the mail? Write a letter to my Congressperson? Please. I can’t even get my neighbors to keep it down after midnight on weekdays. Call me selfish, but I’ve got better things to do than becoming the next John Connor. I might be a Christian, but a crusader I’m not.
So… what of bossmanham’s remark that Luke’s inventing his own ethics? On the one hand, can we really fault Luke or any atheist in that regard? I don’t think so. After all, atheists don’t believe in God, so it’s only logical that they would delegate morality to themselves. On the other hand, I’m less-than-impressed with Luke’s approach to the problem. For example:
…the single largest impact you can have with your charity dollars is to give all of them toward ensuring we develop artificial superintelligence that is friendly to human goals. Giving to stop global warming looks like a drop in a puddle in comparison. [Luke Muehlhauser]
Really? What about all the poor and starving, here and now? Note Luke said the “single largest impact” is to give “all” of one’s dollars towards his charity of choice. I see a glaring discrepancy between this claim and Luke’s stated penchant for empirical evidence. The claim strikes me as pure intuition, quite ironic coming from someone who promotes distrust thereof. Luke’s is, after all, a truth-claim about the real world, and he used to gas on ad nauseum about the need to support such claims with empirical evidence. Has he forgotten the very same standards he used to wage his assault on theists just last year? I don’t get it. What’s even more perplexing is that nobody seems the least bit concerned. I feel like a lunatic when I comment on Luke’s blog these days. In fact, I would wager that I feel just like an atheist stuck in the middle of the Bible belt. Ex?
Furthermore, Luke’s claim strikes me as speciesist. I suspect Luke would disagree, but his sole criterion for artificial superintelligence was that it be friendly to human goals. Why human goals? Albert Einstein is purported to have said that we can’t solve problems by using the same kind of thinking we used when we created them. I tend to agree. Is it not undeniably true that the wanton pursuit of human goals has brought us to the brink of distinction? I mean, we haven’t even begun to arrive at any sort of consensus on human ethics. Isn’t it putting the cart before the horse to jump to machine ethics? If we haven’t come to any consensus about how humans should act, on what grounds should we be deciding how machines should act?
Wouldn’t it be prudent to put this whole thing on pause for a few–hundred years, that is?
Matt
says...I wouldn’t worry about the robot army. I bet they’re just interested in selling action figures. They’re so toyetic!
toryninja
says...Luke seems to have gone off the deep end. He was really good in the begining like you said. But now he just assumes everything he says is settled and will never be challenged again except by ignorant hicks.
Crude
says...Actually, I’m kind of glad Luke’s moved on. Really, once you start to say things like that quote on charity dollars, it just highlights what’s lurking underneath the atheist community. And I was not impressed at his show at Tom Gilson’s blog.
Let me ask you this though: Is it really right to consider someone with this attitude and beliefs an atheist? Exactly how many degrees of difference are there between “we need to direct all our charitable efforts towards making sure the coming artificial superminds that shall lord over us all will favor us rather than be hostile” and “we need to ramp up the sacrifices to almighty Zeus, lest he spoil our crops and disease our livestock”? Or even “I need at least five million dollars or the Lord Almighty will visit punishment upon us all”?
Transhumanists and singularitarians just don’t strike me as atheists. They’re pagans, or at least very unusual theists who are theists all the same. (Look at Frank Tipler and David Deutsch with regards to the Omega Point.)
woodchuck64
says...On the AI emphasis, this post cleared it up for me just now: http://commonsenseatheism.com/?p=15036. When you stop believing in supernatural all-powerful gods, you might start believing in natural all-powerful gods– AIs. I say that with a touch of irony being an atheist myself, but indeed, a friendly all-powerful AI does appear to be an inevitable goal of our ceaseless desires to improve technology for our own well-being.
Matt
says...The concern does make sense. In theory you can make an AI that has the ability to make itself smarter. Once that happens the cat it out of the bag and we are left begging for mercy from our digital overlords.
Read Harlan Ellison’s I Have No Mouth and I Must Scream if you want to imagine a worst-case scenario of non-friendly advanced AI. Of course, your response depends on whether or not you see superior AI as inevitable. If it is, I can see why funding friendly AI research is considered very important. Sort of a Pascal’s wager. Wagering temporary gain against infinite loss.
E.G
says...Cl, i was and am still impressed by your contribution on CSA. You surely know how to keep them humble. Keep up the good work, and i hope that one day you’ll be able to post in your blog as frequently as Luke used to, or as Vox Day keeps doing. Your writings are really good food for thoughts.
Crude
says...If someone believes in a “natural all-powerful god”, they are not an atheist. They’re just a different variety of theist. Really, the god doesn’t even need to be all-powerful, unless people who worshiped Ares and Artemis and such were actually atheists.
I repeat: If you believe in the inevitable coming of an “all-powerful AI”, you’re not an atheist. You’re some variety of theist, on the order of a person who believes in Zeus, or quite possibly the Mormon God.
As for “all-powerful AIs”, does anyone see the problem with trying to second-guess an AI, the purpose of whose existence is at least in part “it will think of things we more limited beings are incapable of thinking of”? Isn’t this at least partially as rational as both A) Planning to build the most powerful calculator possible, and B) Trying to be better at math than the calculator?
With that in mind, I wonder why one option seems to typically be absent from these conversations: “Let’s not build super-powerful AIs.”
cl
says...Hi all. Thanks for sharing.
Matt,
LOL @ toyetic.
toryninja,
I think it’s unfortunate that his faith in the work he’s given has led to overconfidence so great that he actually called God’s existence and dualism “settled issues” in favor of his beliefs. I honestly never expected that, and I know that quite a few others are equally disappointed in that regard.
Crude,
I think that’s a really good question. My initial reaction was to resist the claim that belief in “all-powerful AI” precludes atheism, but the more I think about it, the less sure I am. I guess it all depends on whether one wants to force a non-physical, contingent definition of the word god. Can a god be contingent? Can a god be material? If one answers “yes,” then “AI-ism” seems to qualify as a form of theism.
Honestly, the ones I’ve experienced strike me as Scientologists, not so much in analogy of belief, but in the attitude with which they promote their beliefs. I’m sure there are “down to earth” transhumanists and singularitarians out there, though.
woodchuck64,
Now that sure leads to an interesting line of inquiry… Luke rejects value monism, and claims there are more reasons for action than increase in well-being. Yet, why would AI researchers be on the course they’re on if not to increase well-being? Why would Luke suggest giving charity dollars if not to increase well-being? Of course, these are just variants of this question Luke chose not to answer, and instead deferred to the podcast. If I had a magic genie that could grant me one request in the blogosphere, it would be for Luke to answer that question. At least, that’s how I feel at the moment.
E.G,
Thanks for the good words, although, do you really think I keep them humble? It seems to me that the more I comment, the more arrogant many of them get. I mean, just look at this thread.
Thanks again. I’ve made efforts to post more frequently, but they just don’t seem to stick. I think if I settled for shorter posts more frequently, that might help. Your encouragement definitely helps. I’d like to be at the point where I post something each weekday. I’ve managed to hit that stride before, but, it dies out. It really shouldn’t be a problem. I comment enough on CSA and other blogs that I could easily turn out a post a day, if not more.
Alright, that’s it… [cl rolls up sleeves, indicating work to do] ;)
Crude
says...CL,
“I think that’s a really good question. My initial reaction was to resist the claim that belief in “all-powerful AI” precludes atheism, but the more I think about it, the less sure I am. I guess it all depends on whether one wants to force a non-physical, contingent definition of the word god. Can a god be contingent? Can a god be material? If one answers “yes,” then “AI-ism” seems to qualify as a form of theism.”
Well, we only have to look at the old pagan religions for that. Zeus, Thor, etc were clearly material and contingent. The mormon God is material, I believe, but I’m not sure if the mormon consensus is that God is contingent.
Still, I think if one accepts that those gods were in fact gods, then it becomes pretty suggestive of theism in at least Luke’s case. It’s just an odd kind of theism, some kind of paganism.
cl
says...Crude,
I tend to agree, even though something still doesn’t seem quite right for me. Something tells me Luke would never accept this. I’ve been trying to think of a response that could justify calling oneself an atheist under these conditions, but, I haven’t yet. I would expect Luke to say something like, “there’s a difference between superintelligent machines and gods.” I suspect his reply would hinge on the fact that we create AI, but not the Olympians. If so, the question becomes, “if they are superintelligent, can human creations be legitimately referred to as gods?” To those who say no, I would love to know why not. I’m leaning towards no, but I cannot explain why, at least, not just yet. Must “gods” refer only to entities humans did not create?
I was thinking that polytheism fit the bill: a digital pantheon.
Crude
says...Must “gods” refer only to entities humans did not create?
Well, three twists there.
First, what about idolatry?
Second, granted there will be at least some differences. But then, there’s differences between Zeus and Thor, and Zeus and the God of Western monotheism, etc. I wonder what the salient difference is. Nick Bostrom, when musing about the simulation argument, seemed to explicitly entertain the notion that any programmer(s) of our (assumed simulated) universe would be deities.
A third issue that would get very sticky is revisiting the question of ‘In what we do ‘we’ create, given materialism and determinism?’
I admit that Luke would probably resist being identified as a theist. But then, so what? If he reasonably is one, then he reasonably is one.
(To briefly touch on the original post again: I also think there’s one additional lesson that could be taken for Luke’s case. I’m an advocate of answering atheistic arguments, but not sticking around trying to convince Richard Dawkins he’s wrong. Sometimes all you can do is say “this is settled, on we go” rather than treat the situation as a perpetual argument. I think theists have more warrant to say “the New Atheism failed – here are the counterarguments, here’s the evidence. Now, let’s move on to more pressing matters” generally, even if some individually choose to continue the focus.)
mister k
says...Surely the difference is that most atheists tend to argue that the complex arises from the simple- from the singularity arose the universe, from quarks rose atoms from which arose elements from which arose stars from which arose planets from which arose life. Life creating complex life seems utterly plausible, but the contention is that has to be the order it occurs in. We can’t start with complexity and produce simplicity. Theres surely a massive difference between a belief that some intelligent entity with powers beyond our ken created us, and that we might be able to create an entity with said powers.
cl
says...Crude,
That’s interesting. Wouldn’t that make us the gods, in this case?
Yeah, I was thinking about that, too. It seems to me that given materialism, determinism, and this universe, the ascent of AI is non-contingent. That might be a stretch of the word non-contingent, but the main point is to say that given those three things, AI couldn’t have not arisen. This is basically the determinist’s argument handed back to them.
mister k,
Welcome.
I think that comes the closest to articulating my hesitancy here, although, I don’t see why said entities cannot rightly be considered gods simply because the universe made them.
Rufus
says...Hey,
I thought you all might be interested in this:
http://www.youtube.com/watch?v=CLy0tTfw8i0&
Ray Kurzweil believes that the singularity is not a religion, though it may achieve the aims that religion has sought, i.e. immortality, ending human suffering, world peace, etc.. Ray Kurzweil argues that the singularity is not a religion because religion is “pre-scientific” (which I don’t find much more helpful than simply stipulating the problem away).
What I think Kurzweil misses is that singualrity theory is based on a series of non-scientific metaphysical presuppositions regarding consciousness, identity, individuation, existence, causation, etc. Programing computers with the ability to pass any Turing test is a scientific quest that can be undertaken without such metaphysical assumptions. I can believe that computer programmers will successfully create a machine that would appear intelligent without assuming that the machine is conscious.
But religion ain’t metaphysics, right? Still, if one has an unjustified belief in a metaphysical position, then I am not sure how one does not have a faith-commitment.
So I see the singularity movement as motivated and unified by a series of unjustified faith positions by which they believe that humanity will be saved by a race of advanced immortal, perfectly moral, powerful, enormously intelligent entities. Is that a religious belief system? Kinda? Are the members asking for donations? Hard to define these things… Is Buddhism a religion? Was Spinoza an atheist? Ehhh… Define what you mean by those terms.
Best,
“Rufus”
Crude
says...cl,
That’s interesting. Wouldn’t that make us the gods, in this case?
Perhaps lesser members of the pantheon. ;) Though that leaves the question of who the programmers are, unasked. And Luke there is speculating about AIs being the ones (or one?) really in charge. Really, his
Mister K,
A few problems.
1) The greek pantheon were themselves spawned ultimately from primordial chaos, yet gods they were. That seems like a case of the complex rising from the simple – does that mean the greek’s religion (and many others) were atheistic and naturalistic? Either way, there no longer seems to be that “massive difference” you speak of, at least related to many/most polytheistic gods.
2) Classical theists assert that God is not complex – He is simple. And the universe, in turn, is regarded as complex. So I don’t think your rundown of atheists works here (or at least, ‘the complex must arise from the simple’ isn’t atheistic in and of itself.)
3) You say that we “can’t start with complexity and produce simplicity”, but there’s actually a curious problem here on common atheist worldviews. Namely this: If the universe is eternal – if there is an infinite succession of past moments – then the question of “What came first – the simple or the complex?” gets upended. And one thing is for certain: The complex certainly can produce the simple.
4) Consider the particular difficulties raised in 3 combined with the possibilities raised by people who expect some kind of post-singularity future dominated by AIs, simulations of universes, etc. I’ll frame it this way: If the universe is eternal, then it seems to follow that there have been an infinite number of powerful beings. Remember the sort of powers many transhumanists and post-singularitarians ascribe to future technology: Omega points, creating simulated worlds, flat out creating new universes (John Gribbin), etc. Bostrom’s simulation argument had as part of its core the idea that if humanity goes on to create numerous simulated universes, then we should expect ourselves to be living in a simulated universe ourselves.
That’s me getting off-topic a bit, so to clarify again: No, I don’t think there’s a “massive difference” here. Really, unless quite a lot of actual ‘gods’ throughout history weren’t really gods at all, then I think it’s clear Luke and those like Luke are just some odd form of theist. (And if we’re going to talk about massive differences – surely there’s a massive difference between atheists of the past who claimed to believe in neither Zeus nor the Zeus-like, and someone who claims that Zeus and the Zeus-like will be created, and if we aren’t careful Zeus will punish us rather than reward us.)
Rufus
says...Hey all,
I wanted to clarify my earlier post, if I can. The question of whether “singularitarianism” posits belief in “gods” forces us to examine how language backs our intuitions. Since no such religion existed prior to “singularitarianism”, we might be tempted to either stipulate a definition of religion which fits, or does not fit. This is what I accuse Kurzweil of doing. He provides a stipulation, which only reinforces the position he wants to take. The charitable person will grant the stipulation insofar as it is a personal stipulation and no debate over theory occurs.
On the other hand, we can consider whether this new case, i.e. “singularitarianism,” intuitively meets with our vague notion of “religion.” At best, we can only really discuss how similar “singularitarianism” is to the term “religion” as it is commonly used in our language. We might say that they are relatively similar and that further investigation may warrant a less-than-personal stipulation by which the definition may become accepted in common usage.
This happens all the time. I believe Carnap uses the example of fish. Prior to biological stipulations, the English language referred to animals that live in water as fish. Thus, we find Hermann Melville refer to Moby Dick as a great fish despite is careful, though antiquated, cetology. The modern reader protests and even young children recoil at the perceived mistake in this classic. Whales are mammals! But Melville was appealing to the use of his language and there was no reason at the time to make such sharp distinctions. In effect, Melville was right to call whales fishes, they were at the time! As the biological sciences advanced, the need arose for more precision in our animal classifications. Thus, scientific stipulations led to a turn-over of the lexical definition.
Likewise, an analysis of “singularitarianism” with regard to its metaphysical, epistemological, moral, and political commitments, may lead to further similarities between “singularitarianism” and “religion”. If so, then stipulation may lead to a similar lexical shift, which would warrant identifying adherents, like Luke, theists.
As of right now, I think the weight of the evidence leans towards identifying it as relatively similar to religion, so I find the stipulation not utterly arbitrary.
Do you disagree with my linguistic analysis, or with my conclusion that the weight of evidence leans towards the two being similar enough to define one in terms of the other?
I hope this was clear.
-Rufus
Crude
says...Rufus,
I don’t have a very deep comment right now, except to say I’m not sure it’s true to say that “no such religion” existed prior to “singularitarianism”. Part of what I’m advancing here isn’t that all the god-AI talk is new, but really, it bears some striking similarities to some age-old religious beliefs.
I think this goes double when talk of the singularity, or the (Tipler/Deustch) Omega point, etc, gets painted as “inevitable”. In fact, I will note that – for what it’s worth – the wikipedia entry of “eschatology” seems to have a small section dedicated to singularity beliefs.
Rufus
says...Crude,
“I don’t have a very deep comment right now, except to say I’m not sure it’s true to say that “no such religion” existed prior to ‘singularitarianism’.”
I suppose I was a bit ambiguous there. I mean only that there is no previous religion that is exactly the same as “singularitarianism” and that it has features that seem similar to other religions and features that lie outside of other traditional religions. The question is whether the differences are relevant and whether their are sufficient similarities to justify predicating one of the other.
Thanks for drawing out this point of un-clarity, I don’t think we disagree generally.
This “inevitability” is a striking feature of “singularity” thought. What I have noticed on “Less Wrong” and on Luke’s blog is a general sense in which major philosophical issues are settled and that we must move on quickly to do the real work. But, the motivation behind describing these positions as settled is dubious precisely because it is by settling these beliefs that the urgency is felt in the first place. For all their uber-rationalism, I am surprised by this question-begging methodology. Major epistemological, metaphysical, metaethical, and ethical issues are declared settled because it is urgency felt that no more time should be wasted navel-gazing. They want to head the singularity off at the pass before it becomes “un-friendly.”
That this is question-begging is not even to mention the fact that most seem to be committed to varieties of utilitarianism which are potentially dangerous. It makes me uncomfortable to think that they think they have the theory down and it is only a matter of writing the algorithm. If they are right about everything but the ethics, then they might be responsible for bringing about an even more dangerous version of the singularity than if they had proceeded with caution.
I’ve also noticed a strong anti-philosophical tone within “Less Wrong.” Eliezer Yudkowsky seems to claim that philosophy has little to contribute and that his community can figure it all out on their own. In doing so, they shut themselves off from potential critiques. Luke has his foot in both worlds, but based on my readings of his post, he has already shut himself off to what he perceives as “bad” philosophy, i.e. philosophy that does not reinforce his belief that friendly AI is a pressing concern.
Just some thoughts… back to work.
-Rufus
cl
says...Rufus,
Excellent analogy. This point just refueld my interest in the discussion, big time. Now, we’re under no compulsion to judge “AI-ism” by a rigid, non-robust definition of religion. While we certainly don’t want to “define our way to victory,” I think you’ve raised a very valid point. We might need to parse out in more detail what makes a religion a religion. We haven’t even begun to discuss things like the attitude towards the gods. If we worship AI in a similar way people worship traditional gods–perhaps by trusting their superintelligence by faith, asking them to reveal things to us, or what have you–that has to factor into the discussion. That is to say, at least part of the definition of religion would seemingly have to address the mentality of the adherents.
Crude,
Good observation. “AI-ism” seems to contain it’s own eschatology, and–making it even more similar to traditional religion–there is disagreement on the who-what-why-where-when!
mister k
says...Crude, I don’t believe the definition of god as both simple and intelligent is defensible. If one retreats to deism one can get away with it, as the form of god as something that started the universe is getting pretty close to an atheistic hypothesis for all intents and purposes (its there to start the universe, but then has no affect on our attitudes or our beliefs after that).
A scientific hypothesis that produces the complex from the simple does require a mechanism for that- evolution being the classic one, which is where the Greek gods don’t really have one.
Before I get to my main point, I think we need to be careful about definitions here. When you mean theism you seem to be implying not what most people would think of as theism- a belief in deities that in some way shape our existence- but instead the belief that supernatural entities exist.
But the difference between the theist and the atheist here is not that the atheist rejects the very possibility of the powerful, the extreme intelligence, but the atheist claims that thus far there is no evidence that said beings exist. The atheist might be convinced that said beings are theoretically creatable, but bear in mind that an AI that can eliminate humanity is nowhere near a being that can reshape a universe, the latter requiring not only an intelligence explosion, but for laws of the universe being internally shapable.
I strongly disagree with the notion of an eternal universe. I think its all ill formed concept. We know that the universe has been in existence for 14 billion (or so, thats a number I snatched from the air, but its really not important), but for all intents and purposes we don’t know if the concept of time is even useful before then. Still, one could be led towards a simulation type hypothesis, if one wasn’t cautious. i’m not convinced by such arguments, because it seems I’m about to get mugged by ludicrous probalities (note that the universe simulating me wouldn’t actually have to obey the rules of the universe it creates- we can simulate universes that are drastically different to ours already). I don’t think arguing about such a thing without empirical evidence is useful, because thought experiments can get us going round in circles.
You might then counter that theres exactly the same problem with AI. To an extent you are correct- I think some of the suppositions that Luke and Less Wrong make are based on a chains of “suppose thats”, but there is a core of emperical evidence- we’re making smarter computers all the time, and there seems to be nothing inherent to the problem that will stop us from continuing to progress.
bossmanham
says...Sorry I haven’t commented yet, CL. I haven’t had much time or motivation for blogging lately.
I think my biggest surprise, though it shouldn’t be, is that people think that at some point we’re going to have to extend ethics to machines when there’s no reason to think they’ll ever be anything than unconscious matter that may at some point mimic human behavior. But that’s what you get when you think we’re nothing but biologically machines, I suppose.
I just think that these atheistic ethical theories are exercises in futility and become relativistic at their core.
cl
says...I intend to respond to the last comments from Rufus and mister k in more depth at time allows. For now, I’m wondering: can a legitimate religion be founded on another human being? Let’s say God–the ruler of the spirit world, the Creator of all–doesn’t exist, but Jesus Christ does. Everything flows from evolution.
Is Christianity still a religion? If yes, then why couldn’t the faithful following of some other product of evolution also be a religion?
cl
says...Hey bossmanham. I didn’t mean to leave you out of my last comment. You got yours in while I was posting mine. I’ll get back to you later as well. As for now, “real” life calls, if there is such a thing :)
mister k
says...Cl: you’re right, but a religion is characterised (usually) by a belief that the central thing being worshipped is a good thing. Its certainly possible for belief in a product of evolution to be a religion,or cult- one could make such an argument about scientology for instance. I don’t believe that the singularity meets those standards, but there is a clear similarity between the modes of belief. A belief in a future coming event that humanity must prepare, with only marginal evidence that it will happen, does seem slightly religious in nature.
Crude
says...Mister K,
I don’t see what deism has to do with the doctrine of simplicity. Anyway, on we go.
A scientific hypothesis that produces the complex from the simple does require a mechanism for that- evolution being the classic one, which is where the Greek gods don’t really have one.
Why would they need one? The issue is that the greek gods were by your own standards a case of the complex coming from the simple. Sure, they didn’t have a fully worked out scientific theory of how. But then, there’s a lot of things we don’t know about evolution, or programming. Really, a pretty casual way of describing the origins of the greek gods and demigods would be this: Abiogenesis and evolution.
When you mean theism you seem to be implying not what most people would think of as theism- a belief in deities that in some way shape our existence- but instead the belief that supernatural entities exist.
I don’t buy this. Numerous atheists themselves constantly draw a direct line of comparison between the God of Christianity and Zeus. Not to mention the coming super-AI / Omega Point “shapes our existence” as well. In the case of Frank Tipler, he eventually went the whole 9 yards and started equating it with the God of the OT, though he’s very fringe there.
But the difference between the theist and the atheist here is not that the atheist rejects the very possibility of the powerful, the extreme intelligence, but the atheist claims that thus far there is no evidence that said beings exist. The atheist might be convinced that said beings are theoretically creatable, but bear in mind that an AI that can eliminate humanity is nowhere near a being that can reshape a universe, the latter requiring not only an intelligence explosion, but for laws of the universe being internally shapable.
First, I disagree with your estimation of atheists – I think they’re broader than that. Some say ‘there’s no evidence’. Some say ‘there’s insufficient evidence’. Others say ‘there’s reasonable evidence, but also as much or roughly equal evidence for the opposite’. Atheists are all over the map. And frankly, I keep running into atheists who, when pressed, end up being deists after all.
Either way, Zeus couldn’t shape the universe. Neither could Ares. Neither could many other polytheistic gods. And yet, gods they were. All this before noting that the gods of the singularity are also touted to be beings capable of making exact simulations of our existence, as I noted. In which case we have to start wondering, “Well, then is this experience I’m having right now part of a simulation?”
I strongly disagree with the notion of an eternal universe.
Alright, but many atheists disagree with you.
You might then counter that theres exactly the same problem with AI. To an extent you are correct- I think some of the suppositions that Luke and Less Wrong make are based on a chains of “suppose thats”, but there is a core of emperical evidence- we’re making smarter computers all the time, and there seems to be nothing inherent to the problem that will stop us from continuing to progress.
Some small amount of empirical evidence is easy to come by – that’s a very low bar.
Let me just stress this point once again: Look at Zeus. Zeus was a powerful, but limited being. He had tremendous power, but not dominion over the universe. He was not eternal, but came into being from what ultimately was the formless universe, and he wasn’t even the first god out. There are many gods like Zeus in religious history.
With this part of the singularitarian wing, we’ve got people providing so certain a view of the far future that it seems reasonable to classify it as an eschatology. They are frantically talking about how to secure the ethics of a coming being who frankly is described in ways that suggest more power than many gods of the past, and doing so based on a tower of what-ifs.
Let me throw this one at you. If we live in a simulated universe, does it follow that some form of theism is true?
mister k
says...Yes, it does. I’m a little inclined to say no, because I think that the word “theism” comes with it attached lots of notions of worship and a belief that entities constrain our lives in some manner, but certainly a simulator would be an all powerful god to us. The simulator would exist outside our universe, and be able to make drastic changes to us (at least turn it off again, one would suppose). Simulation requires intelligence somewhere, so yes, a simulationist is a form of theist, but not one that most people would recognise (a belief that one is living in a simultion is extremely unlikely to alter ones day to day activities, unless you claim far more knowledge about said simulation than you are privy to).
Even with this looser definition, I think you are basically making the definition of theism useless if you apply it to the belief that beings with extraordinary power might be creatable by human beings.
I think theres two things going on here
1-Theres an undercurrent in this argument that one might go “you say you’re an atheist, but you believe in the singularity, which makes you just as much a theist as I am”, which just isn’t true, and is something I’d argue against.
2-You could argue that by accepting the singularity Luke is making intuitive intellectual leaps with very small amounts of emperical evidence, and thus is proving no more rational than the theists he contends are wrong.
I feel like argument 2 is more solid ground, although I don’t full agree. I think Luke is being as irrational as say, a deist or a simulationist, but not as someone who believes in any human god.
cl
says...If I might interject for a second…
mister k wrote,
What I’m interested in is, what makes that not true? The only real difference I’ve seen so far is that the “gods” of the singulartity flow from us, whereas a “flesh and bones” god or a spiritual god exists concurrent and/or independent of us.
Is this really a substantial enough difference to not call “singularitarians” theists? I mean, if 97% of a definition applies… ?
Rufus
says...cl,
“What I’m interested in is, what makes that not true? The only real difference I’ve seen so far is that the “gods” of the singularity flow from us, whereas a “flesh and bones” god or a spiritual god exists concurrent and/or independent of us.
Is this really a substantial enough difference to not call “singularitarians” theists? I mean, if 97% of a definition applies… ?”
This is a significant divergence from classical theism. However, if the proposal is that singularity belief is akin to Greek polytheism, then I do not think this is so different. The Olympians were brought about by more primordial, but presumably less powerful “gods”. Also, some of the gods in the Greek and Roman pantheon were little more than heroic humans (http://en.wikipedia.org/wiki/List_of_Greek_mythological_figures#Deified_mortals), which suggests that the existence of a god can involve humans as proximate causes. I don’t know if there is a particular example of a human creating something God-like in Greek mythology. Mary Shelley’s subtitle to Frankenstein, “The Modern Prometheus” comes to mind as a modern myth of just such an attempt. We also have HAL 9000 in 2001: A Space Odyssey. I wonder how influential science-fiction has been on the world-view of many of these singularity believers. Would they admit such influences?
-Rufus
mister k
says...The difference is that a super AI would be created by well understood natural processes. Now we don’t know quite what these are yet, but they’ll involve some of the things we know on building AI already, synthesised with other fields. If an AI becomes self perpetuatingly intelligent then it will have done so using natural methods, breaking no known laws of physics. Gods are always postulated to have been created by some unknown means, their causes being supernatural in their nature. Sometimes its claimed thats its actually impossible to observe these methods.
Crude
says...Some quick comments in passing.
The difference is that a super AI would be created by well understood natural processes.
More about this below, but for now: Isn’t that a bit of a cop-out, considering one of the driving reasons to create the super AI is so it could understand and do things we more limited creatures can’t? That’s actually one (distinct from this conversation) reason I find part of the “let’s make sure it’s a good AI not an evil AI” conversation baffling: Let’s try to second guess the thing for who our motive for creating it is largely “it can think of things we can’t”.
One of the major points of creating these super AI is that they’ll be able to think up things humans are incapable of. Not ‘haven’t gotten around to yet’ but are physically incapable of doing as well.
And, how about a thought experiment: Tomorrow, a massive computer housing a super-powerful singularity AI falls from the sky. We didn’t make it. It uses a technology that’s downright alien to us. We do not know how it works. Is it a god?
Gods are always postulated to have been created by some unknown means, their causes being supernatural in their nature. Sometimes its claimed thats its actually impossible to observe these methods.
I think you’re working off a lot of casual conversation and assumption here, not reality. What was “supernatural” about the origin of Zeus? Remember: Zeus is just some powerful guy who did powerful things. He was quite clearly physical. He originated out of earlier beings, who in turn originated out of chaos. Sure, the ideas were (apparently) wrong and sound quaint now. So do phlogistons, miasma theory, and Lamarckism (to degree). That doesn’t make them supernatural.
I think you’re mistaking the fact that people reflexively call anything that has to do with gods (especially age old gods) “supernatural” with actually being supernatural, and further confusing that designation as meaning much of anything. If I were using the standard of the God of Aristotle, or of Western Monotheism, you’d have a great point. That God created nature, period. He was immaterial, without part, pure goodness, etc. He was not “created by some unknown means” because He wasn’t created, period.
But that’s not the God I’m talking about. I’m going by Zeus, etc. Now, maybe someone can turn around and say “Zeus and company weren’t gods after all”, but man, that’d wreak some havoc on our understanding of things.
mister k
says...I suppose you’re right, I think I keep losing track of what exactly I’m arguing here. I think the argument I just made actually speaks more against argument 2. My argument against 1, the notion that Luke is a theist is rejectable because of type of belief. Luke believes that its possible (not sure how possible, I’m not sure he’s made that clear) that there could be beings created with more power than we can currently possess. Having a belief that such beings could be in existence on another world, or brought into existence here, is different to believing that said entities exist on earth- all evidence suggests that they currently do not after all.
Of course you can make arguments that Luke’s reasons for believing that these entities can be created are similar in many ways to theists, and I might be inclined to agree, but its not quite the same argument.