Why we're the only intelligent life in our galaxy

[DemoCoder]

(note: you can't prove something to be absolutely true, but if you accept as an axiom that physical laws are immutable, you can prove something to be true within some degree of precision).

Not really. Even if we assume that "physical laws" do not vary with time, and even if we only care about provable accuracy, there always remains the possibility of a "special case" that you can't eliminate.

You need a couple of additional axioms, such as that physical laws do not vary with reference frame and observer, so that where you conduct the experiment, and whose's doing it don't represent a special case.

Even then, I'm not so sure. Imagine that the truly underlying nature of the universe is governed by a type of planck level cellular automata. This automata gives rise to emergent macroscopic observables like momentum and position, that (lucky us) happen to be model-able using simple analytic mathematical equations, and can predict the future output of this cellular automaton with reason accuracy.

But what if the output of this automaton is only predictable because it happens to be periodic and we are in a "stable" region of its output right now, one that happens to be "compressible" into physical laws. What if, under certain rare conditions, it produces a "blip", and blips are not predictable by any model (inherently random) and cause experiment to disagree with theory sometimes, but not others.

We're then left with a scientific theory that works, say, 98% of the time to 1 part in 10 million accuracy. But 2% of the time, it fails to agree with data for some reason.

One way you can view this, is that the cellular automata is not time/space invariant and violates your first axiom. My view is that the automata is invariant, it doesn't change its rules, it just happens to be unpredictable.

I think that our confidence in a scientific theory increases everytime they survive an experimental test, and that the alternatives fail. In principle, this confidence can continue to increase without limit, but it will never be 100%, and I think I can live with that.

Somewhat technical point, but it is rather doubtful that we'll be capable of the computing precision and speed required for such a prediction before it can be done in a lab.

I somewhat agree. Abiogensis starts with a simple molecular soup and studies how more complex organic molecules can form over time. It's actually a simpler problem than say, protein folding, because the molecules you are modeling are alot simpler than a protein.

My idea of how this might work is that simulation searches identify candidate "soup" mixtures and settings which produce probable organic formation. Then lab work takes these search parameters, and conducts physical experiments. Data from physical experiments are fed back into simulations to fine-tune and visualize what's happening. and on and on.

 

[DemoCoder]

Yes, I agree. But apart from entertainment, development falls broadly into life science or automation nowadays.

I disagree. People are still building space probes and space ships today (X-Prize anyone) Automation will only create more surplus wealth to devote to such endeavors. There are a myriad of inventions out there that have little to do with life extension or industrial automation. Hell, just go look at the Military budget and what they're ineventing. Does ABL fit into either of your categories?

What do you call the Internet?

 

[Frank]

Current biological machines (life on earth) are dependent on an energy process which would be unsustainable in harsh environments of space. Life based on an artificial substrate, whose information storage can be copied and backed up and transmitted, would be much more robust over eons. Thus it is most likely that interstellar colonists are artificial lifeforms.

We humans won't settle the Galaxy, our *children* will (our robot children). I disagree that their values will be completely divorced from humanity. For sure, they will be different, because non-scarcity and immortality will evolve away many of the evolutionary psychological traits we have developed. However, they will most likely inherent what is best in humanity, our traits which are not dependent on competition for scarce resources in a valley or village. They will be our children, and we will impart to them (by design) many of our values. They may eventually evolve out of it, or reprogram themselves, but I don't think they would neccessarily have completely 'alien' values. They may in fact have values we consider saintly, as such machines may lack alot of the selfishness we have. Such machines may not care about self sacrifice due to immortality, and may not covet other's resources due to non-scarcity. They may very well be machines of pure love.

Would they care about us, or regard us as raw resources to be consumed? That would assume that such machines would need resources so bad that they would override any consideration of *what* resources they consume (or who), but with an abundance of time and power, would they need to be greedy? One would then have to assume the 'anthill' scenario where they eventually don't even consider on their plane of existence. But do non-abusive human beings hurt animals on purpose, or by accident? I don't go around kicking over anthills. I do it by accident, but I actually feel sorry for the ants sometimes watching them rebuild a destroyed house. We are then left with the idea that hyperintelligent beings would fsck us over by accident, not realizing they are stepping on us. I also tend towards this being unlikely.

Thus, it is conceivable that artificial nearly timeless lifeforms colonizing the galaxy might have some respect for lower life forms.

Yes.



I don't think you can throw their motives in with the notion of moral values in this respect, but that could go either way and I agree that it is much less likely that they will be violent than benevolent in general, no matter what race designed their initial form and rules. But we'll have to wait and see, don't we?

Pity it won't happen in our time, though. But you can't have everything.

 

[Frank]

More precisely, the two theories are incompatible, and one or the other, if not both, must be incorrect on some level.

This doesn't mean they're not useful, of course: the calculations which one can do are accurate enough within their resective realms of applicability that they will be useful theories a long time from now, even after we've found new, more exact theories. The theories are probably still quite a bit more accurate than the degree of precision to which they've been proven to date (note: you can't prove something to be absolutely true, but if you accept as an axiom that physical laws are immutable, you can prove something to be true within some degree of precision).

And that is also about evolution: the evolution of ideas into useable tools. And, as we've discussed in length, it becomes harder and harder to come up with something that might be the next step, unless it can encompass all that evolution (experimental evidence) in the first try, or be completely mathematical and artificial in nature (strings).

We need a new paradigm, which might require a new kind of evolution to go beyond the current scientific hierarchy (because that's what that is about) and be able to fit it all together. And a lot of patience to jump through all the hoops and do all the experiments.

In other words: we have to automate all that. Nicely on target, isn't it?

 

[Frank]

I disagree. People are still building space probes and space ships today (X-Prize anyone) Automation will only create more surplus wealth to devote to such endeavors. There are a myriad of inventions out there that have little to do with life extension or industrial automation. Hell, just go look at the Military budget and what they're ineventing. Does ABL fit into either of your categories?

What do you call the Internet?

Fortunately, yes. But look at the volume!

Safe is better, seems to be the paradigm for this day and age, even for scientific research.

 

[KimB]

We need a new paradigm, which might require a new kind of evolution to go beyond the current scientific hierarchy (because that's what that is about) and be able to fit it all together. And a lot of patience to jump through all the hoops and do all the experiments.

Well, the is the purpose of pretty much all of science. Some of it comes in at attempting to grasp at the fundamental laws that govern our universe (SUSY, string theory), but most of it comes down to applying fundamental laws that are known to some degree of precision (Newtonian dynamics, General Relativity, QED, QCD, etc.), and discovering how these laws govern the behavior of complex systems.

The latter is what experiments based upon the origin of life are all about, as well as other experiments in solid state physics (superconductors, magnetoresistance, nanomaterials, etc.), astrophysics (planets, stars, galaxies, galaxy clusters, etc.), as well as any science done in any field other than physics, which I am not really qualified to comment specifically upon.

 

[Frank]

Yes, that's why we call it evolution: it's a slow and gradual process. Which is exactly why it is starting to fail: even if you can answer all the questions to satisfaction, how would you go about getting your new theory that surpasses quantum electrodynamics and relativity in one go accepted? Can it be done, by human beings?

 

[DemoCoder]

You won't be able to automate it without AI, and even then, it will be very limited. Hilbert dreamed of something much simpler: machine verification of theorems. It was shown that this is impossible in general, and not just because you could construct Godel statements. It also violates the Halting Theorem. (and more fundamentally, in algorithm information theory, you can directly show how any AI with a finite set of axioms cannot verify theorems of sufficient entropy)

Thus, if your idea is of a computer which generates theoretical models and then tries to show that the new theoretical model reduces to the current ones in limiting circumstances (like how Newtonian dynamics is a subset of GR), it may be that a general algorithm to do this is impossible.

Humans can only do this (and only do it in limited circumstances, they still can't prove undecidable statements) because they apply special case reasoning to each scenario. Therefore, your superduper automated AI would have to simulate and evolve thousands of different theorem provers as well. And then you'd have to worry about the correctness of those theorem provers. We don't worry about human correctness because we know humans are flawed, and we rely on peer review. But who is the peer reviewer for the millions of theoretical models and provers being spun out? If the machine theorem provers are wrong, some true models could be eliminated, or you could be spammed with millions of bogus resultsf it's human beings, you haven't automated anything, since we have to do the grudge work of verifying everything the machine is doing.

It must be other machines then. Machines which evolve. Pretty soon you end up with a race of AI having to have evolved just to do our theoretical work for us.

 

[NANOTEC]

(all quotes from the debate I've paraphrased)

What, you don't think the question, "What is the one piece of evidence that could absolutely disprove your whole idea?" was relevant?

The ID debater flatly refused to answer the question the first time it was posed.

The second time it was posed, it answered with a statement similar to:
"If somebody could prove that life could have arrived accidentally, then I will resign my position."

Later in the debate, the debater against ID says:
"I believe that within a decade we will have produced artificial life. ... What scientists will do is they will produce the conditions that existed in the early universe. If they do this, and create life, will that state that life could possibly have arisen accidentally?"

To which the ID debater answers "No."

Like I said, this is nothing more than dogma on the ID side.

If you listen carefully the second question was more rhetorical than sincere. That is why the ID guy responds with his own question which was "who will do the sequencing?".

 

[Tahir2]

What scientists will do is they will produce the conditions that existed in the early universe.

So what you are saying is that they will design parameters into the experiment that they believe were like the early Universe, (which will probably later on be shown to be incorrect anyway by other scientists), and then life will sprout up?

Why not get a bit of soil, add water and sunlight with a few sprinklings of Cress seeds and let's find out what will happen. Great experiment.

 

[KimB]

If you listen carefully the second question was more rhetorical than sincere. That is why the ID guy responds with his own question which was "who will do the sequencing?".

And while the use of vocabulary may be questionable, I'll stick with what I think his intent was with that question, and answer with nobody (which was the responder's answer as well). The entire point of such an experiment would be to see whether or not, given the right conditions, life can be generated spontaneously (or, more accurately, what specific conditions are required, as any scientist worth his salt working in the field of the origins of life believes wholeheartedly that it can be generated spontaneously).

There would be no point at all to the experiment if an experimenter manually tweaked the reactions.

 

[KimB]

So what you are saying is that they will design parameters into the experiment that they believe were like the early Universe, (which will probably later on be shown to be incorrect anyway by other scientists), and then life will sprout up?

Well, in the young Earth (sorry about the misstatement).

If the material makeup was shown to be incorrect, then it would certainly invalidate the original experiment, and the scientists would have to do more work with the new makeup (though such an invalidation is unlikely: there's no "one" environment on the Earth, there are many, and there always have been).

After all, we've already done the first parts of such an experiment. They were done many years ago, in fact. The entire inspiration for deciding what sorts of materials should be placed into the soup to see if life could be born was studies of the young Earth's makeup. These things are very tightly-related.

But this is why those of us in science often believe so strongly in some of our basic theories: They're not just supported by one piece of evidence, but by many. When we settle on what we believe was the environment that gave rise to life on Earth, it must be supported by many completely independent bodies of evidence to be so accepted. For example, you expect that if you recreate the right conditions in a lab, you should be capable of producing all of the building blocks of life (I am currently a bit skeptical as to the anti-ID debater's statement that we will create life in this way: I suspect the probabilities are small enough that it'd take much longer than the experimenters have time to wait, but we should be able to find a set of circumstances which create all of the essential components within a short time frame and a similar or identical environment). Secondly, you should be able to do terrestrial experiments to show that such an environment once existed on Earth, to the last atom in the equation. In the very long term, we should also be able to do experiments that show that any location with this same environment, over a long enough span of time, also gave rise to life.

Why not get a bit of soil, add water and sunlight with a few sprinklings of Cress seeds and let's find out what will happen. Great experiment.

What you are stating here is out of complete and utter ignorance of how scientists go about their experiments, and how those experiments are investigated by other scientists.

 

[Tahir2]

So what you are saying the experiment needs to be done in such a way that spontaneously life occurs where there was none before.

If the experiment is trying to reproduce conditions with an aim in mind then we have our first problem.. was the Universe actually like this in its early stages.

The second problem is that if the Universe was like this and life has sprouted, it still doesnt answer the question of why. Why was the Universe like this that it meant life sprouted.
If you cannot answer the why, the how is boring and inconsquential.

It's like saying here is the Universe before (blank page) and now we are looking at how the ink is written on it, is it textured, does it ripple, does it spontaneously appear and disappear. Also ask where did the paper come from and why was the paper there in the first place? Then lets talk about what ink has got to do with the origins of the Universe and if it can actually tell us where it came from. Unless of course ink spontaneously appears out of nothing too... then the question is WHY does it appear spontaneously... one slip of the finger and all the formula's stop working to explain life, the universe and everything else. Why?

 

[Tahir2]

..complete and utter ignorance of how scientists go about their experiments,

Thanks mate.

You also forget that you showed extreme bias and had concluded like all scientists "worth their salt" that life spontaneously appears. Not good for starting point for science and an experiment.

 

[KimB]

The second problem is that if the Universe was like this and life has sprouted, it still doesnt answer the question of why. Why was the Universe like this that it meant life sprouted.
If you cannot answer the why, the how is boring and inconsquential.

The why is a religious or philosophical question, and unimportant to asking the how. Which is more interesting is totally a matter of personal opinion. I claim that anything that is unknown is going to be interesting to somebody, and there's a whole hell of a lot about how the world works and how things came to be that we don't know, things which are therefore very interesting to some.

It's like saying here is the Universe before (blank page) and now we are looking at how the ink is written on it, is it textured, does it ripple, does it spontaneously appear and disappear.

It's much more than just that. We're looking out a small window into the Universe and asking how things got to the way they are today. That's a very hard question to answer, and requires lots of experimentation as well as theoretical work.

 

[DemoCoder]

Tahir you are confusing two totally separate questions. 1) how and why does Abiogenesis happen. 2) Why are the the parameters of the natural physical laws such that Abiogensis is possible. Are these parameters unique, or is it the case that there are many variations suitable for life-like complexity. That's the anthropic principle question.

The answer to question #1 certainly isn't boring. It would involve being able to determine, from basic physics and chemistry, what is it that enables self-assembly and spontaneous order to arise.

The second question involves another "why". Of course, you are delving into religion, and my RELDAR "Religious-Radar" is going off, because there is an infinite descent of "Why" For me, the idea that God created it all is boring and uninteresting. It explains *nothing*. Why did he do it? What is the nature of God? Who created him? Are there hierarchies of Gods? etc ad infinitum.

Part of the problem is that there always isn't a "Why". See Gregory Chaitin's *proof* (absolute *proof*) that the reason that the vast majority of mathematical statements may be truth or false is inherently random, can't be explained, and there is no "Why"

For example, consider the Twin Prime conjecture. There are an infinite number of twin prime numbers separated by 1 even number. It's never been proven. It's possible that it's nonetheless truth, but CAN'T BE PROVEN. Why is it truth? *NO REASON WHATSOEVER, IT JUST IS* and you may have to assume it as an experimental axiom. Why is the universe the way it is? Perhaps, *IT JUST IS* and there is absolutely NO REASON WHY.

 

[Frank]

You won't be able to automate it without AI, and even then, it will be very limited. Hilbert dreamed of something much simpler: machine verification of theorems. It was shown that this is impossible in general, and not just because you could construct Godel statements. It also violates the Halting Theorem. (and more fundamentally, in algorithm information theory, you can directly show how any AI with a finite set of axioms cannot verify theorems of sufficient entropy)

Thus, if your idea is of a computer which generates theoretical models and then tries to show that the new theoretical model reduces to the current ones in limiting circumstances (like how Newtonian dynamics is a subset of GR), it may be that a general algorithm to do this is impossible.

Humans can only do this (and only do it in limited circumstances, they still can't prove undecidable statements) because they apply special case reasoning to each scenario. Therefore, your superduper automated AI would have to simulate and evolve thousands of different theorem provers as well. And then you'd have to worry about the correctness of those theorem provers. We don't worry about human correctness because we know humans are flawed, and we rely on peer review. But who is the peer reviewer for the millions of theoretical models and provers being spun out? If the machine theorem provers are wrong, some true models could be eliminated, or you could be spammed with millions of bogus resultsf it's human beings, you haven't automated anything, since we have to do the grudge work of verifying everything the machine is doing.

It must be other machines then. Machines which evolve. Pretty soon you end up with a race of AI having to have evolved just to do our theoretical work for us.

Yes and no. If this trend continues, no human being without super persuation powers will be able to get something revolutionair accepted. We're just about there right now. Most new things we see nowadays are political or marketing in nature, like ID. Because it has become easier to force something to acceptance through those channels, than by offering it up as a new scientific theory.

Scary!

And let's be honest here: without funding you're not going to make it. No matter what you came up with.

See it as a case of consolitis, not for PC games, but for the scientific community. Every one of them has to have that budget, and wants to see his or her name as often as possible in credentials. And more than half the battle is shooting down the competition, if you're only deaing in ideas.

Ok, it isn't as black and white as that. But it takes plenty of time for others to comprehend what you came up with, while it's much easier to shoot it down according to common knowledge. Look at what it doesn't do, instead of at what it does.



And while we're not there yet to automate the creative part (AI's, as you said), I think they will happen gradual as well. Look at gene splicers and recombiners, for example. We have better equipment nowadays than pictured in "Jurassic park", for example, while everyone was very sure when that came out, that those things would not happen for at least a hundred years...

Btw, I think creativity is almost equal to "seeing many sides of the issue". Having choice. It's not something inherently random or intuitive. But those might be a product of having a good understanding of all the issues and viewpoints involved. Enough data and the ability to choose the best option for each scenario, in other words.

And that's how it goes. Automate it, build a machine that really does it, and you have won. Seeing is believing, and cannot be denied by people who don't care about "that stupid other thing".

AI will happen, or it won't. But there isn't any definite limit to the amount of things we can automate.

 

[NANOTEC]

If something can't be observed in principle, then it is unscientific. If it can't be observed for technical reasons (your particle accelerator ain't strong enough, your telescope isn't good enough), it becomes a gray area with two possibilities. 1) observation is technically feasible and within reasonable time horizons 2) observation is technically not feasibleThat in essense, is science, and without it, we are shit up the creek. We invented the written word, because "oral history" resulted in in "copying errors" of information, where facts and events are distorted as they are transmitted. And we invented mathematics, because descriptions of phenomena written using words are too error prone, since words mean different things to different people. And we invented rational criticism: peer review of the reproducible, because people make mistakes.

As for explanation, one can only interpret what scientific models say, one can't "prove" an explanation. Quantum Mechanics has half a dozen or more "explanations" and none of them can be proven to "more correct" than the other. They all give exactly the same answers and predictions, so it comes down to aesthetics.

I accept that science has its uses, but I also accept that it has limitations. For example the Theory of Evolution which says all life evolved from single celled organisms. It's a good theory but it cannot fully explain the origin of life. The theory has too many missing pieces. I accept it for what it is an educated "guess".

You suggested it.

No I did not suggest it was started by another life form. I only said ID may be one of the many possibilities.

Give us an example of a non-random natural event.

Sleeping, eating, drinking, deficating.

If we "don't know" anything, than whether it's 10, 10 billion, 100 billion, or a trillion, we don't know what the number means either. You're weaseling now.

I didn't say we don't know anything. I said we don't know enough to say we're likely the only intelligent life in our galaxy. The fact we exist says there is likely others who exist or have existed.

Proof can only be attained in a closed system via deduction. You can't prove anything in science. Science is based on inductive reasoning. Inductive reasoning can't achieve 100% certainly when dealing with the real world. As I said, you can't even prove that the Sun will rise tommorow. How many times does this have to be explained to you?

The fact that you don't "get it" shows that you don't understand rational thinking. It's the same flawed thinking that makes Intelligent Design debaters look like dopes.

And what you don't get is that the Theory of Evolution is one guess among many guesses. Whether it has more supporting evidence than other theories says nothing.

Well, we know there is one: Humans on the Earth. Perhaps you meant two? Why not three? How do you arrive at the idea that the Earth implies that one other intelligent species MUST exist? (your definition of likely)

Otherwise, what is it? 50% 25%?

Whether it is 2, 5, or 9 makes no difference. The point is still the same. We're likely not alone.

No theory can be proven. Whether a theory is accurate or not is a different issue. Quantum Electrodynamics is an extremely accurate theory. It has agreed with experiment to amazing levels of precision. Relativity is so accurate that we can build amazing things like the Global Positioning System. Are they proven? No. We can say whether something is accurate, we just can't say it is proven.

Again, it would help if you went and got a book on the basics of the philosophy of science.

In science you have a hypothesis then you test it. A theory has no defined sample size. The origin of life cannot be tested to show that single celled organisms from the ocean evolved into humans. It only shows how one species of organism can evolve over time.

A science guy who doesn't understand science and bashes it. Wonderful. The religious people will still love you.

Actually I do understand it, I just don't take it as gospel. It has limitations and the Theory of Evolution exposes those limitations.

Well, your attacks on evolution are becoming tired. Science already did a good job of explaining how single cells came to become animals. The people who don't believe in cellular evolution today are frankly loons. Now, Abiogenesis. People have a pretty good idea of some of the components of Abiogenesis and there is certainly a scientific basis for it, for we can observe many organic precursor molecules arising spontaneously today.

When a serious Abiogensis theory is eventually created, it will be testable, as in, we'll be able to create life in the laboratory and show it arising out of nothing. It may be done with real chemicals, or it may be done on a supercomputer like Blue Gene's great granddaughter.

Until that day comes Evolution will still be one of many possibilites.

 

[Frank]

Btw, don't you all think that the possibility of the predominant lifeform having silicon ancestors is much higher than them having, say, human, DNA based, or any other kind of ancestors? Xbdestroya might be right in suggesting silicon life as a good candidate after all.

 

[KimB]

Thanks mate.

You also forget that you showed extreme bias and had concluded like all scientists "worth their salt" that life spontaneously appears. Not good for starting point for science and an experiment.

Yeah, well, it's somewhat hard to convey the culture of science that I've been exposed to in my three years of gratuate school, and two years of undergraduate study at a research university prior to this.

To put it simply, scientists are incredibly skeptical. It takes a lot of convincing to get the majority of the community to move to a new view. It requires multiple experiments performed by completely independent groups just to verify the same set of experimental data. Then it requires measurement of completely separate data sets to convince people of a theory that connects the separate data. And even with all of this convincing, every scientist out there will probably tell you, if you ask the right question, that they don't believe that the theory in question is absolutely correct.

But when you're working on the origins of life, there are some simple assumptions that you must make in order to get any amount of work done. One of those simple assumptions is that all natural laws which we can measure today worked in the same way in the past (note that this in part defines what we mean by natural laws). The entire idea of intelligent design is antithetical to the use of natural laws to describe our world: it states that some intelligence interfered with the natural laws.

Intelligent design is therefore, by the structure of the "theory," not subject to experiment: one cannot, in a lab, manifest the same intelligent designer that is proposed to have started life out here on Earth and get it to do something for you.

Therefore, Intelligent Design must be thrown out as a possibility for performing science. This isn't an argument that it didn't or couldn't happen. This is just a statement that if we want to do any work at all in understanding our universe, we cannot make use of the idea of Intelligent Design.

This doesn't come down to belief, or bias, or personal preference. It is pure and simple logic: if we want to make use of logic to discover the nature of the universe, we need to make certain assumptions, and one of those is that everything can be described through a set of (possibly not yet known) immutable natural laws.

Note to DemoCoder on immutability: Stating that currently-known laws may change with time is merely a statement that there is some higher framework which describes this change, and this higher frame work does not change in time.