Fine-Tuning?

[Zarathustra]

Well, I'm finally done. I'll have more to say later, in fact, I think I might be starting a few threads about it. Probably in the Tank. But for now, I just have to say this is one of the most important books I've ever read. I've never seen such a comprehensive world view from epistemology to physics to math to art to politics to ... reality. This helps me sort some of my ideas, giving me a more cohesive framework for many things I've already been thinking. [For instance, his attitude towards global warming and prophetic pessimism in general are exactly what I've thought for a long time.]

Thank you, Peter, for this recommendation. No one I know in Real Life has done me such an intellectual favor.

[I've checked out your blog, btw. You're holding back on us, man! Your passion would be perfect for the Tank. ]

[peter]

Thank you, Peter, for this recommendation. No one I know in Real Life has done me such an intellectual favor.

That is praise that I shall value for a looong time. I'm so glad it was a good call; recommending books is a risk fraught enterprise at the best of times.

[peter]

I'm going to flip back here from the 'beauty' thread because I re-read the last chapter of Deutsch last night in the knowledge that there were some points I wanted to raise once Z. had finished.

I was very taken with [what I refered to earlier as a 'curler'] Deutsch's summing up comments in respect of advances that have been made in cosmology since the publication of his previous work. His account of how, in an expanding followed by collapsing universe the formation of 'omega-point' universes allows for the speeding up of time such that in the final analysis the universe still remains infinite in duration was an eye-opener [we will be headed toward the 'big-crunch' but will never effectively [from our perspective] get there], but this I think paled into (almost) insignificance in respect of his next observations. He refered to the entity we call 'dark energy' and was unequivocal in stating that we haven't got a clue as to what it is. Never the less, it seems that it's presence is the only explanation for the volta-face that has occured in cosmology that now indicates that the universe is infinitely expanding.

This in turn would appear to negate the need for the paralell universes that the fine-tuning explanation rests on in the anthropic explanation. The infinite vista of time that an ever expanding universe allows for would allow of itself [if I have got this right] for all of the things that are demanded by the paralell universes of the collpsing/expanding/collapsing model. Ok - thats fascinating - but then Deutsch gets into weird terratory. He proposes a thought experiment - no - I'll come back to that. He says that it is likely that in the future, the entire universe will be simulated repeatedly in it's entirety for study purposes. On this basis the philosopher Nick Bostrom has speculated that it is far more likely that we now are experiecing a simulation than actually existing in the real thing. Deutsch refutes this simulation argument by proposing a thought experiment in which you envisage space as actually being many layered like 'puff-pastry' in a french patissary cake. These lakers split and re-join as do we with them, but we are oblivious to it [nb all the layers are identical.] Deutsch spends time in London and also in Oxford; now say London has a million layers but Oxford only one, then if he wakes one day in a dark room and has forgotten whether he is in London or Oxford should he bet that the overwhelming likelyhood is that he is in London since he is instantiated a million times more often in Lonon than Oxford. He feels not. He feels that the instances of the layered time do not equate to 'histories' and that it is the histories that are significant. Using this thought experiment he refutes [wrong word maybe] the simulation argument presumably in favor of a 'fifty-fifty' likelyhood of us being in the real universe as opposed to a simulation.

It was a very thought provoking final chapter - not least because of Deutsch's unflinching admissions as to what we do and don't know. He didn't seem to think we were all that close to understanding what is needed to produce a viable AI, he stated flatly that the two basic models underpinning our understanding of 'how the world works' are wrong and we know it. So contrary to what many great minds have previously 'prophesied', in respect of our journey toward the forming of good explanations, we really still are at the beginning of infinity and whats more we always will be.

[Zarathustra]

I've always thought that the people who think AI is just a matter of making computers fast enough, large enough, are naïve. The problem is much harder, namely, knowing what intelligence is in the first place. It will take understanding creativity itself.

But Deutsch thinks we'll be able to do this, though we've underestimated the problem. However, this is the one part where I'm tempted to apply Penrose's thoughts that consciousness isn't algorithmic. Deutsch points out that reducing human choice to an algorithm misses the possibility of creating new options which aren't spelled out in the algorithm. Okay, that's where creativity comes in. But then does this mean creativity can't be rational? No, he still thinks we can explain it and then code for it. But isn't that a contradiction in the idea that choice involves options outside of any potential algorithm? Or does he think that the creative act itself is the creation of "higher order" algorithms? Kind of like how humans can see the Incompleteness of Gödel's theorem by moving up to the "next level" and seeing the incompleteness of the previous level? But then if you can program an algorithm that creates higher order algorithms, you're once again merely stating that consciousness and choice are merely algorithms.

The parallel universes idea isn't necessary to explain fine tuning, at least not the kind of parallel universes suggested by quantum mechanics, i.e. the multiverse, which is the collection of parallel universes whish (he claims) has overwhelming empirical evidence. The idea of an infinite universe (or multiverse) makes the idea of multiple multiverses unnecessary, which are the ones invoked to explain fine tuning.

[peter]

The algorithm/choice arguments you refer to Z. would seem to be problematic in the terms you express them [though I won't pretend I had seen them as such untill you have drawn attention to it]. As you said before however, the mans train of logical thought can be difficult to follow at times and you wonder what explanation he would pose if you put the question to him. He made the observation that while mathematical proofs are rarely found to be 'false' as such, they are however formed into higher levels of arrangement [previousely unknown] and in this sense are [in the same manner as scientific theorems] continually improved upon. This however still leaves the 'algorithm within the algorithm' problem you refer to.

On a different note I was also intrigued by his comments regarding the moral aspects of 'qualia simulation' untill when/if we get a full understanding of what a 'qualia' actually is.

Lastly Z., I think it worth noting that you may indulge yourself in at least a small 'pat on the back' for your having of your own rational thinking, independantly come to that very same optomistic stance that Deutsch has made the central theme of this book [by no means a majority viewpoint] - and it is clear that he is one of the foremost thinkers of the day. You made reference to having read some of the comments in my [alas pretty defunct] blog. I will state without reservation that the reading of this book has caused me to have a serious rethink about many of my views on humanity and our prospects for the future and I think [and hope] I am a better person thereby.

[Hashi Lebwohl]

The other problem with trying to reduce AI to an algorithmic system is that algorithms have to exist in a logical, axiomatic system and Godel, whom you mentioned, tells us that axiomatic systems are by their very nature incomplete. I doubt any sort of AI which can approximate human intelligence can be built without quantum computers, which do not rely solely on 0 and 1.

[Vraith]

The other problem with trying to reduce AI to an algorithmic system is that algorithms have to exist in a logical, axiomatic system and Godel, whom you mentioned, tells us that axiomatic systems are by their very nature incomplete. I doubt any sort of AI which can approximate human intelligence can be built without quantum computers, which do not rely solely on 0 and 1.

A quantum computer would certainly make things go more quickly/efficiently, if it works as expected.
But that doesn't mean it is necessary.
Fuzzy logic [and some other related options and seemingly similar but genetically different systems that I'm not sure have been modified/used to write code yet] overcomes the 0,1 problem. Even though the machine itself is still operating on that basis, the code isn't.
The algorithm problem can be addressed/compensated for [is already done so in a number of areas] by stacked and overlapping code/instructions.

I somewhat agree with what Z said above...it's harder than some folk think...but I don't go all the way with it. I don't think we have to know exactly what intelligence is, nor creativity. I think we'll have a working AI long before we understand those things.
[[which is why/where the AI/Human conflict...if the doom forecasters are correct, which I doubt...would arise from. That understanding gap.]]

I think that for mostly one reason: we humans have a very long history of creating functioning implements/devices before we know the underpinnings.
Hell, a pretty damn powerful argument could be made that in most cases we don't...and maybe CAN'T...understand the underpinnings till AFTER we build a machine.
I mean...for how many tens of thousands of years did we use fire...not just AS fire, but to make OTHER inexplicable-but-working things...before we had any clue what it was/how it worked?
How many billions of people used levers and wheels long before even one person had the math and physics to explain it?
I definitely think we'll create intelligence before we comprehend it.

[peter]

And that is exactly, I think, what makes a lot of people nervous.

[Vraith]

Ok, I bought the thing a while ago, but needed a time for my first reading...for me that is a speed through with as few external [time to feed the cat] and NO internal [what the hell is he talking about?...think think...continue] gaps.

I may revive this thread [cuz it has cool content already] or start another [cuz this one has vectors already] after I do my second read, or third, but so far:

I am more impressed by certain things. [explanation, how it functions].

But less impressed by more. [though I emphasize they are all instinctual, preliminary...heh...conjectural, and I'm not going into detail here/now.]

Here is the primary problem:
No explanation is a "good explanation"---it is only better than the prior explanations.
But then he says objective knowledge is attainable...even though "proof" isn't the point, or even possible in nearly all cases.
A whole mess of stuff is involved with all that...and most of the implications are really freaking COOL as HELL, IMO.

But then he ruins a lot of it by making claims that are false about other views. And calls them "bad philosophy" or "bad explanations," while, at the same time co-opting PART of those things [which is verboten, according to him...no combined/compromised methods/explanations can work]
Just one example, his utter disdain for postmodernism as a whole.
Yet, he USES ideas that were born, evolved, and created good explanations from that frame. [he even, in one of the last few chapters, says so explicitly...and he seems to smart to have not known he was doing so]

Another: his claim on the infinite hotel, where a guest accidentally tosses their puppy in the trash [REALLY? Who does that????....but it is funny anyway] and therefore it vanishes...irretrievably...no, even worse than that. No one, anywhere, can give it a new home...cuz it has somehow become, non-existent and non-existable in the future by any possible means. But not because it is DEAD. That's just normal.
But, what he has done, really, is place time/place/mathematical constraints on the "Infinite Hotel" that inherently VIOLATE the rules that allow the Hotel to exist, even as a mental/conceptual/abstract model or "thought experiment."
[[it is also probably in contradiction with his ideas on creativity...though it's hard to tell, since he doesn't know what it is or where it comes from. That's not his fault though...it's simply a lack of knowledge of everyone so far. But you'd think, since he is so insistent, he'd recognize the quandary].

Despite those things, I'm very glad I read it...no doubt he's smart as hell and has a bunch of "people should pursue that" ideas.

Soon-ish, once I've had some meme-acquiring and testing time, I'll do a critical read. We'll see what comes of that, if anything.

[peter]

Terrific news V.! Now I get to see you and Z. critically discuss a book that has certainly blown me away and I really get the chance to learn something!

The book and any discussion it generates is certainly worth a thread/threads of it's own [IMO] because it's implications go so far beyond the 'fine tuning' strictures of this thread [not that I think we took much notice of them anyway ].

I won't ask too much untill you are ready to comment fully, but on first reading, dou you share Deutsch's almost fundamentalist optimism in respect of the human species and it's place in the Universe. [It's certainly contagious as you read it on the page!]

[Hashi - you've got to be next to read this - your input is way to important to be absent ]

[Vraith]

but on first reading, dou you share Deutsch's almost fundamentalist optimism in respect of the human species and it's place in the Universe.

I pretty much always have. I don't say it that much around here [or IRL either], cuz I have concerns almost all the time about how lots of things are, where they're heading, blah blah blah. But FUNDAMENTALLY, the odds are that we will, as people, survive and far more than survive. Ascend.

One problem I see...and it MAY, heh "derive" from something D talks about...is the eternal, short-sighted, "just-in-the-nick-of-time" thing. There are so many problems that we can see while they are small...but we never even consider dealing with them till they are huge.

This is a "bad" meme. It's a static-society born meme.
And it has a cousin that it just loves...they spend almost all their time together: it blames the CAUSE of the problem on the ideas/structures/methods that could SOLVE the problem. [teen sex/pregnancy/abortion is CAUSED BY sex education]

Another related one is the "We can't AFFORD it." But I'll probably deal with that once I've gone into it more deeply. It's just a conjecture at this point.

Those particularly...and a couple other symbiotic bad ones...are beginning to gain power in the West. [they're not new, the cure/vaccine exists and for a while the West was innoculating. But we're now treating the anti-vax crowd as if they make sense, letting them cause real harm to others.]

And I'm always aware of the chance of failure [so is D]. I THINK if we could displace the only-when-it's-near-disaster meme, a number of other problems/genetically related static memes would vanish. [we won't get rid of problems as a field/effect. But we'll think of/approach them with a different, and better, attitude. D obviously believes that it's the attitude towards, not fact OF problems that matters...and I basically agree.]

So, yeah, I think the odds are highest for good things.
I think I'm just a slight bit too old to live long enough to see the turning point where almost everyone recognizes and believes it. There is a critical mass, I suspect, and I don't know what the exact figure is...where we end up with nearly infinite beginnings of infinities almost all at once.

And it's probably not even a number. It is certain events [and I think I know the most likely ones. I mean both most likely to happen AND most likely to change people's minds/attitudes] Prediction, not prophecy, and only probable...but something I plan to think about further once I start reading critically].

But I also think that we do many many things in the worst way possible. Even when we know better...and we very often do. We kill a lot more people, and let [and MAKE] orders of magnitude more live body-and-soul shattering lives, than is necessary. And we do it, almost always, with eyes wide open and minds shut.

[peter]

One fear I always have is that optimism can morph into denial - or do I mean that denial can be mistaken for optimism. It is one thing for example to be optimistic about mankinds ability to solve our current environmental issues - quite another to deny they exist [or that they are not in part man-made.]

I have used pessimism in the past as a protective cloak against dissapointment, and I see after reading Deutsch that it is a cowards trick and unworthy of a rational thinker, but that does not mean to say I believe in abandoning a healthy degree of 'realthink' in my thinking; you have to be realistic in your assesments of what may or may not happen or the act of thinking at all becomes a fools errand.

But enough. It's enough for me V. that you have embarked on the reading of this book and are going to give your critical opinion of Deutsch's 'thesis'; as Z. said, there is little in life that the book does not touch on in some way. I'm not going to be falsely modest here V. - Deutsch is a way smater guy than me and in so being it makes his [what seem to me on the face of it] highly reasoned arguments difficult to dispute. Already you have mentioned areas where it may be that his logic begins to unravel [though no doubt D himself would dispute it vigorously ] - and I'm not in the 'pride' business, I'm in the learning about how things really are in the world' business, and if D's world view has to be sacrificed on the alter of that cow, then so be it - but it's going to take better heads than mine to do it.

[Vraith]

hey, remember D's infinite hotel thing?
It's real/physical.
Well, kinda...
phys.org/news/2015-10-physicists-experimentally-quantum-hilbert-hotel.html

[peter]

Looks interesting V. Will read in full on my other device coz that pop-up is annoying as **** on this small hand-held!