Hacker News Reader

Comment by chroma | Hacker News Reader

Looking at the paper linked to in the article[1], I'm having a hard time not dismissing this immediately. There are several implications to this theory:

- Information has mass.

- Information cannot exist at absolute zero.

Does this mean that bringing a hard drive to absolute zero changes its mass and erases its contents? Does the information somehow come back after the drive is warmed up? Also there are many ways to represent information: magnetic charges on a spinning platter, electrical charges in SSDs, physical impressions on metal, graphite on paper, etc. Do all of these get destroyed at absolute zero? I don't know how that's reconcilable with the rest of physics.

1. https://aip.scitation.org/doi/10.1063/1.5123794

View on HN

Former mathematical physicist here. I haven't read the article yet so don't take this comment as a defense nor as a rebuttal, but I just want to point out that your reasons for immediate dismissal are not obviously valid.

- The Bekenstein-Hawking black hole entropy directly relates information to entropy. There has been an "it from bit" program (and more recently "it from qubit") dating back decades that tries to treat information as somehow fundamental with matter/energy emergent. The jury is still out, but I wouldn't consider it to be particularly controversial (at least not moreso than other speculative theoretical physics).

- To me it seems absolutely physically plausible that cooling a hard drive to absolute zero would destroy its contents. As another example, you can destroy the information contained in DNA at much warmer temperatures than absolute zero. Why would heating it back up restore the information? Most thermodynamic processes are irreversible. And by the way, thermodynamic irreversibility is related to entropy change, which is a measure of information lost.

> To me it seems absolutely physically plausible that cooling a hard drive to absolute zero would destroy its contents.

Cool, cool. It doesn't, but enjoy.

Why would you imagine this would happen? The magnetic domain is not affected thus, and neither are MOSFETs.

You're a mathematical physicist? You're the single least likely person on Earth in my imagine to say "oh, you reduced the speed of some atoms? That must mean magic happens."

What could possibly wipe a hard drive from the action of cooling it down?

> Why would heating it back up restore the information?

In short form "it doesn't."

You may be confusing that some heating processes create damage with the idea that warming something up will harm it.

As any schoolchild can tell you, DNA has survived billions of years on this planet, being heated and cooled somewhat rapidly throughout.

> And by the way, thermodynamic irreversibility is related to entropy change, which is a measure of information lost.

This is a very "consider a spherical cow" approach to physics.

The reason physicists are made fun of on those grounds is that the abstractions they bring to the discussion are frequently so far divorced from the real world situation that by the time you start bringing the real world back to the discussion, their abstractions fall apart.

Real world objects that were not designed for durability have not suffered this damage on the timescale of a fifth of the length of the universe.

That things are irreversible and related to one another isn't much use when you look at the real world and the things you're describing aren't actually happening at large scale, even in uncontrolled natural conditions.

An insect that's been in amber for 500 million years typically has DNA loss on the order of ~5%. It has been exposed to a wild range of temperatures.

If you genuinely believe that exposing a hard drive to absolute zero would destroy it, help us understand why Voyager is still running.

Maybe it's because it's still a tenth of a degree kelvin above absolute, or something?

There's a huge difference between exactly zero and close to it. Space isn't even that close compared to what's been done in a lab, and even in lab experiments the temperatures involved don't correspond to such effects happening.

Cool story.

Did you have an actual explained mechanism by which this is going to happen?

Because if you're saying "exactly zero and close to it," we've never actually had a single atom at exactly zero, let alone a macroscopic object

We can get close enough to absolute zero that very weird quantum effects being to happen which one could argue are associated with information loss.

Cooper pairs is one such example, where at very low temperatures electrons pair up and begin to behave like a new combined particle with integer spin. Thus they morph from fermions to bosons which are no longer subject to pauli exclusion and can all occupy the same state. The information required to describe such a system decreases substantially as the potential state space is now quite limited.

This manifests in physical phenomenon like superconductivity and superfluidity.

We cannot do these things macroscopically, no.

I will repeat my protest which you ignored from the previous comment.

1) We have never reduced even a single atom to absolute zero. Practicioners debate whether it's actually possible.

2) Your discussion of bose-einstein condensates is neat and all, but the explicit context is a hard drive. Nobody has ever condensed a macroscopic object (the breathless article about condensing a tardigrade isn't actually correct.)

There has never been a superfluid or superconductive hard drive.

Please focus on the question being asked, in the context being asked, if you must reply.

How do you propose to reduce a hard drive to actual non-almost zero degrees? Not a few particles, not electron pairs. A hard drive.

Once it's at absolute zero, formally, so what? Yes, I saw you guys handwaving "spooky stuff happens," but in reality, we've had hard drives within a tenth of a degree kelvin and nothing happened.

If you're going to propose effects, please have a specific mechanism in hand that creates the specific effects asked about, in a context that is the hard drive and not two electrons

Thanks

The CERN CMS magnet is a 10,000 ton superconductor. 6-meter diameter and 13-meters long.

For the vast majority of materials (of any size), strange things do happen when you hit the critical temperature. You can take any amount of mercury for example and when it hits 4.1 K it loses all its electrical resistance.

Hard drives are not the only items that can contain information.

A third time:

1. We've had hard drives within a tenth of a degree of kelvin, and the effect you're asserting did not occur

2. You're offering no explanation for what would actually cause this

3. The weight of a random superconductor is an irrelevant detail added for fake technical acumen

We have had these devices at these temperatures, and the thing you're talking about did not occur.

Until you can say why it would, there is no reason to take this seriously.

We've already done this. Your speculation is invalidated by experience and data.

You seem very arrogant discussing things that are purely theoretical (absolute zero objects). A healthy debate includes space for speculation, and you are shooting this speculation down with arguments like ”it just doesn’t work like that”, which does not bring anything to the discussion.

Sure, sure. And I speculate a wizard did it.

When people are speculating that effects will happen, even though we've done this and they didn't, by asserting that an impossible goal wasn't reached, it's appropriate to ask them how.

If you feel that it's appropriate to use emotive words like "arrogant" when someone says "please tell me how that's possible," then I guess I'm not that worried about your opinion.

Indeed, it's a purely theoretical scenario, and I don't think anyone is claiming otherwise: the paper simply notes that one consequence of the theory is that information cannot exist at absolute zero, and then someone who considers this weird questions what would happen to an object storing data were this to happen. Your response was 'nothing, because data storage on space probes still works', which is a non-sequitar: the theory does not predict that data storage at space temperatures will cease to hold information.

But one specific mechanism exists: at a certain point there's entropy in simply the structure of the macroscopic object which prevents lowering its temperature further. So if you were to cool it as close to absolute zero as possible it would require turning it into something which was not a hard drive (again, this is something would occur far below the current achieved temperatures with even microscopic objects).

> Indeed, it's a purely theoretical scenario

Theory says that this scenario cannot occur, so it's actually not a theoretical scenario.

Temperature is defined as the brownian net motion of a bag of particles. No reference point is involved.

How do you make that zero temperature? Are these particles zero-motion with respect to themselves? Then they aren't with respect to the planet, or the sun, or the galaxy.

Are they zero-motion with respect to the sun? Then they're hot enough to melt in seconds.

The conceptual idea of zero-temperature is not a real thing.

These people doing their fake-wise "but weird things happen at zero temperature" are just snake oil salesmen plying false knowledge.

There is no such thing as zero temperature.

In the meantime, there's nothing theoretical about the practical scenario, either. Voyager's hard drives have been in the absolute zero of outer space for more than half a century. Its heater, which did not reach the hard drives, has been off for more than a decade.

Those hard drives have been sub-1-kelvin for years and nothing changed.

These people are trying to speculate about what would happen if in a situation we've already had a dozen times, because they don't know the truth, and are filling their lack of knowledge with guesswork.

When it's presented to them that facts exist, they attempt to move the goalpost to theoretical limits which physics says aren't actually real at all, and then from there try to lean on their depth in physics.

> So if you were to cool it as close to absolute zero as possible it would require turning it into something which was not a hard drive

This is a fantastic point, and something I hadn't even thought of.

I like this a lot.

I'm still wondering about "what, you think ball bearings in a vacuum don't work when it's chilly?"

Like I can't even think of why they think it'll fail, except some magical belief that the fact that it's cold just causes magical breakage

The operation of the damn drive is friction based. It'll heat up! Crimeny.

I'm mostly done with this conversation, but I'll note that space is not as cold as you're thinking: The coldest that voyager is gonna get in the next million years is 2.7K (thanks to the cosmic background radiation this is the equilibrium temperature of any object in deep space), not sub-1K, and it's a lot warmer than that now, latest report is about 200K (though that's about 8 years ago). It also doesn't have hard drives, though this is a pedantic point because it does use another form of magnetic storage. You're talking about a 'practical scenario' which is so far away from the theory being discussed it's completely and utterly irrelevant, like if you were talking about the temperature of the sun it'd make little difference in how much hotter you're talking relative to the relevant temperatures.

Armchair interest only, so please tell me if I'm wrong or at least we theorize how I could be wrong.

My understanding is that even in the full vacuum of space, we do not get to absolute zero. Quantum fluctuations keep it ever so slightly above it.

So perhaps information does actually not exist at absolute zero?

Well I understand that information, like energy, cannot be destroyed (hence Hawking radiation) so it couldn't just be reduced to zero, it'd have to be transformed. Also the parent's argument about freezing and then reheating to destroy and restore information is a good one because that would be reversing entropy in a closed system (assuming the info was indeed destroyed) and thus impossible.

IANA physicist though, just a lay person with an interest in information theory.

According to the paper, the amount of information that can be stored decreases as one gets closer to absolute zero. Our current information storage technologies and refrigeration technologies don't let us actually test this, but with a high enough information density and a low enough temperature, one should expect to lose information if this theory were true.

> My understanding

Correct. The current zero point energy is believed to be a false vacuum (local minimum). One of the end-of-the-universe scenarios involve it tunelling to a lower state (bubble nucleation). Things become progressively more dire the lower the energy of the next (false or real) vaccum is, including matter and gravity ceasing to exist. Not only does this sound a lot like "no information at absolute zero," it is also a terrifying existential crisis (you're welcome).

An atom cannot be at absolute zero, it is an unreachable extreme and all sorts of weird quantum effects go on as you try harder and harder.

The connection between information entropy and thermodynamics entropy is pretty well established, as well as theories which relate the destruction of information to irreversability. This part of the article is actually not particularly controversial. The idea of a storage device gaining some mass when information in placed on it is not particularly new, but it is something which would be extremely difficult to measure. In terms of the question of absolute zero, you have to consider that the physical reality of a system at zero temperature (truely zero, not femtokelvin) and thus zero entropy is extremely weird: it would basically require a perfect crystal lattice extending infinitely in all directions. This is one of the reasons why it can't be reached. So the answer is basically that if you were to cool a hard disk to absolute zero (which is impossible), it would first require turning it into something not recognisable as a hard disk. (And if you're thinking we've come close to absolute zero in experiments, remember that the difference in scales between thermodynamic information/entropy and other information (even the information contained in the shape of a solid object, let alone any amount of information our current technology could store) is incredibly vast: the temperatures reached in labs don't really come close to that.

So while this is still basically a theoretical idea (and likely will stay that way for some time without a very clever experimental design and a lot of resources: notice they basically propose building a LIGO to perform their experiment), it's not as weird as you might think, and the ways in which it is apparently weird reflect a weirdness that is already present in thermodynamics.

As far as I understand it nothing can really ever be at absolute zero, it can only be asymptotically approached.

Reply on news.ycombinator.com