Comment by krapp | Hacker News Reader

Comment by krapp | Hacker News Reader

Parent Full thread

I think people intuitively perceive the universe as being roughly analogous to Earth's oceans. Obviously we know it isn't objectively true, but subjectively we're just not capable of comprehending the true scale and emptiness of space, of reconciling the Hubble Deep Field image[0] with the fact that less than 5% of the entire universe is physical matter. Even the distance between the Earth and Moon is mind-boggling on human scales, and humans have traveled it.

So it seems reasonable that if any interstellar civilization were out there, one would have found us by now, or their presence would be obvious to us. After all, Rome knew of China, even if they never directly interacted, and Spain eventually stumbled across the New World. Von Neumann probes make sense because it doesn't occur to us that the chances of any arbitrary path along the universe intersecting with a star would be effectively zero. We look up at the night sky and we can see so many of them, and they seem so close together, obviously you'd run into one no matter which way you went, sooner or later. Right?

It's more comforting to make up rationales as to why we might be alone than deal with the ramifications of being surrounded by life we may never see in a universe in which the entirety of physical reality amounts to a rounding error in a zero.

[0]https://en.wikipedia.org/wiki/Hubble_Deep_Field

[1]https://www.theatlantic.com/technology/archive/2014/12/the-c...

True, the one thing I keep trying to remind myself and other of is that we really don't appreciate the cosmic scale all that well.

There's a point where it's all just "big". A million and a billion seem equally big. Even though, as it has been said before, that the difference between 1 billion and 1 million is about 1 billion.

There are so many fundamentally difficult problems with deep space exploration, that it may never be logistically possible to over come them. And that's of the ones we know about.

Rome and China were roughly at the same technological level, more or less.

Given modern technology, we would be able to observe both without being noticed. It would be pretty trivial from space. It wouldn't be quite so trivial close up, but it would be possible.

A time difference of ten thousand years? The invisibility would be rather more effortless, and might even include direct memory manipulation in the victims.

A million years? Can anyone even imagine what a million year difference would look like? Never mind the technology - we can't even imagine what its motivations would be.

So assuming an advanced civ is going to do what we'd do - build ships, colonise, hoard and transform resources - is more than a little naive and unimaginative.

It's not unlike a protohominid assuming that life is about hunting, and any other two legged creature they meet is going to want to talk about flint knapping and these new arrowhead shapes.

> So assuming an advanced civ is going to do what we'd do - build ships, colonise, hoard and transform resources - is more than a little naive and unimaginative.

The assumption is weaker than you suggest. It's not that your typical advanced civ will do what we'd do, just that some proportion (perhaps extremely small) of civilizations will be expansionary.

But what is to make them expansionary. Space is so vast and resources plentiful that it is doubtful any run of the mill system would be interesting. After you are sufficiently advanced you probably do not need much resources at all.

> After you are sufficiently advanced you probably do not need much resources at all.

Any process that involves replication will eventually need more resources.

We're not capable of comprehending the true scale/emptiness of space, sure - but it seems like many commentators are having difficulty comprehending the length of time the universe has been around for.

If a single civilization in our galaxy made a machine capable of replicating itself and going to other star systems less than a billion years ago, we should have seen that.

> it doesn't occur to us that the chances of any arbitrary path along the universe intersecting with a star would be effectively zero

I fail to see how this is a strong critique of Von Neumann probes, maybe you can clarify.

>I fail to see how this is a strong critique of Von Neumann probes, maybe you can clarify.

Allow me to attempt to clarify.

Let's say a civilization creates Von Neumann probes. Ignore whether such devices are even technically feasible. It sends these Von Neumann probes out into the universe in all directions.

The chances of any one of them ever encountering anything larger than a grain of dust before the heat death of the universe is practically zero. In almost all possible universes, all of the probes are swallowed up by the void. In almost all of the rest, one probe finds enough material to be able to replicate, and all of those are swallowed up into the void.

The scenario whereby a stable chain of exponential reproduction and growth via interstellar transfer occurs lasting long enough for a significant amount of stellar matter to be consumed and for an obvious trace to be left which just happens to be visible to us, is not the most likely scenario in a universe where such devices exist, it is still so unlikely that its feasibility borders on requiring sorcery. Because the universe is just that empty.

It's the sort of idea that only works on graph paper assuming a "perfectly spherical cow in a frictionless vacuum" kind of universe. The lack of a universe awash in Von Neumann probes is not a strong argument against the existence of life or advanced civilization in the universe, it is, itself, a strong argument against the feasibility of Von Neumann probes.

Why are you assuming that such a probe is not targeted? Targeting radically changes the results. An interesting middle ground to explore is untargeted probes taking gravity into account

Targeted probes would be more complex, and thus more prone to mechanical or software failure, or copy failure across generations. Untargeted probes could remain dormant in interstellar space, whereas targeted probes would have to remain active (and thus consume energy) in order to course-correct and continue to approach their target.

If the goal is to mine the resources of one specific location, then a targeted probe makes sense. But if the goal is to proliferate and explore (or consume) as much of the universe as possible as efficiently as possible, then targeted probes seem less likely to succeed over the scale of time necessary.

> Targeted probes would be more complex, and thus more prone to mechanical or software failure, or copy failure across generations

I'd argue that your argument is either

a. not intrinsically true of complex probes or

b. just a modified version of the great filter argument

You could imagine an advanced civilization that creates an artificially intelligent probe that is pretty effective at problem solving and resilient to these sorts of failures. Even if targeted probes have to remain active, there is no reason they couldn't consume energy harvested from their origin world to course correct.

If your claim is that every form of intelligence will inevitably fall to mechanical or software failure, etc. then that claim seems to just be identical to the great filter claim.

> targeted probes seem less likely to succeed over the scale of time necessary.

Your own GP seems to provide a very compelling reason why this is not true.

Is "targeting" really a large requirement to add on to a device that needs to be able to fabricate spacecraft from raw materials wherever it lands?

>But if the goal is to proliferate and explore (or consume) as much of the universe as possible as efficiently as possible, then targeted probes seem less likely to succeed over the scale of time necessary.

Less likely than non targeted ones that you say wouldn't get past the first generation?

> It sends these Von Neumann probes out into the universe in all directions. The chances of any one of them ever encountering anything larger than a grain of dust before the heat death of the universe is practically zero.

Haha okay. I think most would view this as a pretty uncharitable assumption (appears contingent on these civilizations not having telescopes or AI), but if that's your view I can understand why we disagree.

Sorry I still don't get what you are saying - wouldn't the probes be sent off to the nearest "interesting object" rather that onto a straight line into nothingness?

> The chances of any one of them ever encountering anything larger than a grain of dust before the heat death

This is why no civilization would randomly fire off Von Neumman probes in random directions. They would understand that this is a problem.

If you want to efficiently use them, you fire them at nearby stars. They will either find materials and restart the process, or they won't (and potentially report this, but it's not required). A non-zero number of them may encounter sufficient material (given how plentiful planets seem to be, that's a fair assumption). Those star systems would now fire off a bunch of probes, at near stars. Rinse and repeat. It's a slow process, initially, but given the exponential nature, it should be surprisingly quick.

"Aiming" at star is not something that's too complex even for us. It might require mid-course corrections to account for errors, but we could build a probe that would conceivably reach a star. Ensuring it would be alive and operational by then is more challenging, but it is an engineering one. Actually entering the star's orbit might be even trickier depending on relative speed. But that's peanuts for a civilization that could build such things.

> The lack of a universe awash in Von Neumann probes is not a strong argument against the existence of life or advanced civilization in the universe, it is, itself, a strong argument against the feasibility of Von Neumann probes.

It is. But that's assuming we can even recognize one.

Reply on news.ycombinator.com