Hacker News Reader

Comment by eloff | Hacker News Reader

If nothing else, the EmDrive has been a great exercise in how hard it is to control all variables in a sensitive experiment where the measured effect is small.

It actually blows my mind how well the LIGO folks have accomplished this to be able to detect gravitational waves.

I really like the effort going into this, despite a positive result being impossible to our knowledge. Those are the high risk high reward experiments with the potential to discover new physics. I still expect it to be experimental error, but there's always that remote chance we'll learn something we didn't expect about the universe.

View on HN

> I really like the effort going into this, despite a positive result being impossible to our knowledge.

Not despite, but rather, because of. The fact that it's impossible but we still keeping seeing a small positive signal is what is driving the effort. Nobody likes an unsolved paradox!

> I still expect it to be experimental error

We might even learn something valuable just by figuring out what the cause of the experimental error is, no?

Almost certainly, but maybe the type of valuable thing NASA scientists and engineers learn 100 times a year rather than once in a century.

If we're lucky. Sometimes it's just a loose cable. https://en.wikipedia.org/wiki/Faster-than-light_neutrino_ano...

This experiment has been repeated though, hasn't it? So an error of that kind is more likely to be systemic, at least.

It seems doing microscale gravity experiments face similar limitations. We can't get good atomic scale experiments to even test most quantum gravity experiments. It's fun to read what's required for even milli-Newton experiments.

Similar to the initial experiments in cold fusion.

Part of the problem with “cold fusion” was the claim that it was fusion, which it pretty clearly wasn’t. If the claim was merely that they’d observed apparently unexplained release of energy, or even that they believed they’d detected fission byproducts down near the limits of their instruments, it would have been more palatable. If they’d decided to release their ‘results’ in a peer-reviewed journal instead of in a press conference where they kept the details secret, they might have been treated better by the community. The problem with Cold Fusion wasn’t that it was bad science—-it was that it wasn’t science.

The Em Drive lacked the snake-oil showmanship of Cold Fusion, so it got treated better. It was also described well enough for the experiment to be replicated, and a few confirmations make it interesting enough to explore.

I think this is most likely to turn out to be experimental error, not a cause a revolution in electromagnetic theory, but it is worthy of the investigation, which Cold Fusion never was.

Even if it violates conservation of momentum, that doesn't really mean anything, since this is all happening on a quantum scale. It would definitely violate physics if you didn't need to apply a constant energy source for this to work, i.e. just one laser beam that spins in there virtually indefinitely while providing thrust. However, that is not the case.

This device presumably requires a reasonable amount of energy comparable to other existing drives. So what might happen here is a potentially unexplained quantum effect. Quantized Inertia seems like a really reasonable idea. Why wouldn't it be quantized? It doesn't even make sense for it not to be quantized.

Anyway, I hope this thing works.

Momentum is still conserved in quantum mechanics. This breaks the laws of physics regardless of whether or not it needs a constant energy source.

It breaks the laws of physics as we know them. But we don't know everything, there is still a remote chance.

Conservation of momentum is so well established by this point, that's it's a very remote chance, and I'd expect it to still be conserved in that case, just through a mechanism we did not anticipate.

I'm just really rooting for new physics here, because I think is been about 70 years since we've had that.

Conservation of momentum is, via Noether's theorem, a mathematically proven consequence of translational symmetry in uniform space, i.e. that physics is not affected by linear movement through uniform space.

The bar for violating that is super high, unless you don't believe in mathematical proof.

Of course the space we occupy is not entirely uniform, due to general relativity, but GR addresses that, and the results are not consistent with em drive results, i.e. momentum is still conserved.

Well, maybe they didn't. How would we ever know?

The problem with LIGO is that is not making the expected number of discoveries, it has, as far as I know, one signal that correlated to something visible with other means, and it is so expensive that there won't be independent confirmation soon.

Don't get me wrong, experiments like LIGO are massive achievements. But from an outsider point of view, that very scale and effort seems to be problematic when it comes to reproducing the results of experiments if there are no immediately useful predictions.

> How would we ever know?

That is why they built two of them in two different locations, one in Washington and one in Louisiana. Only results that are measured by both are considered actual signals.

there are a total of 4 now, I believe.

Where did you hear that? Wikipedia only knows of two.

https://en.wikipedia.org/wiki/LIGO

It would be really weird to build more than two because they are incredibly expensive and the incremental value of the third and fourth ones wouldn't be very high.

With several, wouldn't you be able to localize the waves to particular areas of the sky better? That would probably be useful. Expensive, but useful.

Indeed. See the light green (just LIGO) vs dark green (LIGO + Virgo) areas in this[1] picture.

[1]: https://www.ligo.org/images/GW170817_images/GW170817-frontpg...

There are more graviational wave detectors than just LIGO (2). There's at least GEO600, KAGRA and Virgo.

> that very scale and effort seems to be problematic when it comes to reproducing the results of experiments

First of all, LIGO is 2 observatories so every result out of LIGO was observed by at least two locations.

There is also an independent project called Virgo which is making gravitational wave observations, and the results are being correlated with LIGO. They have made a number of overlapping observations, which further moves us towards reproducibility.

Future observatories are also planned (like LISA), which will be another opportunity to further validate events.

Additionally, one of the gravitational wave observations was corroborated by immediate follow-up observations with normal telescopes [1].

[1]: https://www.as.arizona.edu/discovery-optical-counterpart-gra...

I believe there have been visual/radio detections correlated with LIGO detections. I couldn't find a more authoritative source, but did find at least one reference in a Popular Mechanics[1] article.

[1]: https://www.popularmechanics.com/space/deep-space/a28631/neu...

Better source: https://www.ligo.caltech.edu/page/press-release-gw170817

This is in large part why we consider measurements probabilistically too.

LIGO could be wrong, sure, but that's why it's measured twice. The results fit numerical predictions by general relativity that we think the probability of it being wrong is very low. It might not be, but we could be inside a Boltzmann brain too.

Reply on news.ycombinator.com