Slashdot Mirror

evolution of DNA predates Earth by 6 billion years

This fascinating article makes it look quite possible, if not plausible. Seriously good read, and I am not a biologist.

This is one of DeepMind's recent papers, Playing Atari with Deep Reinforcement Learning [PDF]

Definitely the bottom-up approach. It's almost intractible, but it's still easier than attempting to decompose a metric into constituents. A major issue is that general relativity is not linear, so if you've got two Schwarzschild metrics (which describe a spherical mass alone in the universe, edging towards flat spacetime inifinitely far from the mass) you can't simply add them together and get a new metric describing two spherical masses. In the weak-field regimes this will work OK but it's still only an approximation. Away from the weak field this simply doesn't work. While this does make it harder to attempt to connect up smaller-scale metrics to get to a larger one, it makes it basically impossible to go the other way.

So what we can have are models such as Swiss-cheese models, where the universe is built from LTB (or, increasingly, Szekeres) metrics each stitched onto a region of Robertson-Walker. Two sizes of LTB model spherical local voids and spherical local clusters, while if you use Szekeres you're modelling quasi-spherical voids and clusters, at the cost of computational complexity. Or you can use that complexity of a Szekeres, or related metrics, and build more realistic models of local structure than the LTB allows.

An alternative approach (as in papers by Clifton, Rosquist and Tavakol, http://uk.arxiv.org/abs/1203.6... and http://uk.arxiv.org/abs/1309.2..., or recently by Korzyski, http://arxiv.org/abs/1312.0494) is to literally take a universe composed of an arbitrary number of black holes. Surprisingly, you can solve this system exactly. It's not a viable model of the acual evolution, but it is a very good testbed for whether the type of effect I'm talking about is actually significant. The answers appear to be "A universe composed of a large number of black holes looks like Robertson-Walker" but "It is not clear that the dynamics are Robertson-Walker".

"Should choice of metric be kept maximally free at each step?"

In principle, yes, in practice, no. Without constraints all you'd end up with are the Einstein equations split into different scales, with no simplifications. That's not a realistic approach, much as it would be lovely.

Computationally we're definitely going to head towards numerical cosmology as a subset of numerical relativity. One of the major issues is dynamical range; theoretically, and using current perturbative techniques, we can cover a range of momenta from k->0 (edging towards infinite wavelength) all the way to around k~1 Mpc^{-1}, which is cluster scale. That can then be stitched onto n-body codes which can now map down to kiloparsec scales with a significant overlap with perturbation theory. Attempting to solve the whole problem you have to somehow cover at least that same five or six orders of magnitude, throwing at each step the full weight of GR, which is extremely costly. But it will definitely come, and sooner rather than later.

Definitely the bottom-up approach. It's almost intractible, but it's still easier than attempting to decompose a metric into constituents. A major issue is that general relativity is not linear, so if you've got two Schwarzschild metrics (which describe a spherical mass alone in the universe, edging towards flat spacetime inifinitely far from the mass) you can't simply add them together and get a new metric describing two spherical masses. In the weak-field regimes this will work OK but it's still only an approximation. Away from the weak field this simply doesn't work. While this does make it harder to attempt to connect up smaller-scale metrics to get to a larger one, it makes it basically impossible to go the other way.

So what we can have are models such as Swiss-cheese models, where the universe is built from LTB (or, increasingly, Szekeres) metrics each stitched onto a region of Robertson-Walker. Two sizes of LTB model spherical local voids and spherical local clusters, while if you use Szekeres you're modelling quasi-spherical voids and clusters, at the cost of computational complexity. Or you can use that complexity of a Szekeres, or related metrics, and build more realistic models of local structure than the LTB allows.

An alternative approach (as in papers by Clifton, Rosquist and Tavakol, http://uk.arxiv.org/abs/1203.6... and http://uk.arxiv.org/abs/1309.2..., or recently by Korzyski, http://arxiv.org/abs/1312.0494) is to literally take a universe composed of an arbitrary number of black holes. Surprisingly, you can solve this system exactly. It's not a viable model of the acual evolution, but it is a very good testbed for whether the type of effect I'm talking about is actually significant. The answers appear to be "A universe composed of a large number of black holes looks like Robertson-Walker" but "It is not clear that the dynamics are Robertson-Walker".

"Should choice of metric be kept maximally free at each step?"

In principle, yes, in practice, no. Without constraints all you'd end up with are the Einstein equations split into different scales, with no simplifications. That's not a realistic approach, much as it would be lovely.

Computationally we're definitely going to head towards numerical cosmology as a subset of numerical relativity. One of the major issues is dynamical range; theoretically, and using current perturbative techniques, we can cover a range of momenta from k->0 (edging towards infinite wavelength) all the way to around k~1 Mpc^{-1}, which is cluster scale. That can then be stitched onto n-body codes which can now map down to kiloparsec scales with a significant overlap with perturbation theory. Attempting to solve the whole problem you have to somehow cover at least that same five or six orders of magnitude, throwing at each step the full weight of GR, which is extremely costly. But it will definitely come, and sooner rather than later.

Definitely the bottom-up approach. It's almost intractible, but it's still easier than attempting to decompose a metric into constituents. A major issue is that general relativity is not linear, so if you've got two Schwarzschild metrics (which describe a spherical mass alone in the universe, edging towards flat spacetime inifinitely far from the mass) you can't simply add them together and get a new metric describing two spherical masses. In the weak-field regimes this will work OK but it's still only an approximation. Away from the weak field this simply doesn't work. While this does make it harder to attempt to connect up smaller-scale metrics to get to a larger one, it makes it basically impossible to go the other way.

So what we can have are models such as Swiss-cheese models, where the universe is built from LTB (or, increasingly, Szekeres) metrics each stitched onto a region of Robertson-Walker. Two sizes of LTB model spherical local voids and spherical local clusters, while if you use Szekeres you're modelling quasi-spherical voids and clusters, at the cost of computational complexity. Or you can use that complexity of a Szekeres, or related metrics, and build more realistic models of local structure than the LTB allows.

An alternative approach (as in papers by Clifton, Rosquist and Tavakol, http://uk.arxiv.org/abs/1203.6... and http://uk.arxiv.org/abs/1309.2..., or recently by Korzyski, http://arxiv.org/abs/1312.0494) is to literally take a universe composed of an arbitrary number of black holes. Surprisingly, you can solve this system exactly. It's not a viable model of the acual evolution, but it is a very good testbed for whether the type of effect I'm talking about is actually significant. The answers appear to be "A universe composed of a large number of black holes looks like Robertson-Walker" but "It is not clear that the dynamics are Robertson-Walker".

"Should choice of metric be kept maximally free at each step?"

In principle, yes, in practice, no. Without constraints all you'd end up with are the Einstein equations split into different scales, with no simplifications. That's not a realistic approach, much as it would be lovely.

Computationally we're definitely going to head towards numerical cosmology as a subset of numerical relativity. One of the major issues is dynamical range; theoretically, and using current perturbative techniques, we can cover a range of momenta from k->0 (edging towards infinite wavelength) all the way to around k~1 Mpc^{-1}, which is cluster scale. That can then be stitched onto n-body codes which can now map down to kiloparsec scales with a significant overlap with perturbation theory. Attempting to solve the whole problem you have to somehow cover at least that same five or six orders of magnitude, throwing at each step the full weight of GR, which is extremely costly. But it will definitely come, and sooner rather than later.

For those too lazy to RTFA and comment section,
Karo Michaelian says:
January 22, 2014 at 3:52 pm

The theory for the origin and evolution of life as presented above and accredited to Jeremy England is not new. It was published by myself in 2009, K. Michaelian, arXiv:0907.0042 [physics.gen-ph] http://arxiv.org/abs/0907.0042 and again in 2011, K. Michaelian Earth Syst. Dynam., 2, 37-51, 2011 http://www.earth-syst-dynam.ne... The observation that under a generalized chemical potential material self-organizes into systems which augment the dissipation of that potential should be accredited to Ilya Prigogine, “Introduction to Thermodynamics of Irreversible Processes”, John Wiley Sons Inc., 1968. I have written a number of other papers on the thermodynamic dissipation theory for the origin of life, including an explanation of homochirality. These papers are freely available by searching for my name “Karo Michaelian” on ResearchGate. I welcome Jeremy’s contribution to the effort to understand life from a thermodynamic perspective.

Please look at this paper, as well as to the other papers published by prof. Loinger.

To be accurate, the search in Gran Sasso is a search for WIMPs (Weakly Interacting Massive Particles), which are one microphysical explanation for dark matter. I personally do not like the common conflation of dark matter (for which there is abundant evidence) with WIMPs (for which there is no evidence at all).

A lot of the interest in WIMPs comes from particle physics, due to the "WIMP miracle" (that hypothetical particles at the electro-weak scale, i.e., ~ 100 GeV, apparently have the right mass to explain dark matter) and the hypothesized connection between WIMPs and supersymmetry (i.e., that the WIMP could be a supersymmetric neutralino). After much experimental work, the WIMP miracle is almost dead experimentally, and the supposed connection to supersymmetry is not doing so well either.

However (not that you would know from reading most articles on the subject), there are a number of other viable theories for dark matter. These include axions, primordial black holes (maybe), and macroscopic quark nuggets, which would have important practical implications should they be detected.

To be accurate, the search in Gran Sasso is a search for WIMPs (Weakly Interacting Massive Particles), which are one microphysical explanation for dark matter. I personally do not like the common conflation of dark matter (for which there is abundant evidence) with WIMPs (for which there is no evidence at all).

A lot of the interest in WIMPs comes from particle physics, due to the "WIMP miracle" (that hypothetical particles at the electro-weak scale, i.e., ~ 100 GeV, apparently have the right mass to explain dark matter) and the hypothesized connection between WIMPs and supersymmetry (i.e., that the WIMP could be a supersymmetric neutralino). After much experimental work, the WIMP miracle is almost dead experimentally, and the supposed connection to supersymmetry is not doing so well either.

However (not that you would know from reading most articles on the subject), there are a number of other viable theories for dark matter. These include axions, primordial black holes (maybe), and macroscopic quark nuggets, which would have important practical implications should they be detected.

The actual paper can be found here: http://arxiv.org/abs/1401.5761. People have suggested informally ideas somewhat similar to this one before, but Hawking proposal seems to actually have the math behind it. Possibly most importantly, he can show that his predictions are a consequence of gauge/gravity duality http://en.wikipedia.org/wiki/AdS/CFT_correspondence. This suggests that this may be a testale consequence of certain string theories if one could observe a black hole under the right conditions and see that it only was pretending to be a black hole.

http://arxiv.org/abs/1106.0847

One of the most interesting physics papers I've read in recent years. Does away with dark matter by presuming that antimatter has the opposite gravitational sign as matter (which pops out very naturally once you apply CPT to general relativity).

As the electromagnetic force is almost 10^40 times stronger than gravity, it would be virtually impossible to test with anti-protons or positrons. But with electrically neutral anti-hydrogen, it becomes potentially testable.

Our results for one particular benchmark do not rule out the possibility of speedup for other classes of problems and illustrate that quantum speedup is elusive and can depend on the question posed.

The study asks a very specific question and acknowledge its limited scope.

The Air Force's revelation was based off of previous research into this in real birds in 2008.
http://arxiv.org/abs/0804.3879

Except you know here's something that certainly came out when they were alive.

http://arxiv.org/abs/0804.3879

Sure, it's NP-hard, but it's also nonrecursive via Rice's theorem.

That depends a bit on how the problem is formulated. When stated in full generality, ie "given an arbitrary UTM, decide whether it's a regex finder", it's obviously not decidable.

However, the full expressive power of Turing machines may not be needed. For some classes of grammars/automata, the question whether something is a member of the class can be decided by an automaton, this property is known as automaticity and has origins in the field of descriptive complexity. See for example
http://semukhin.name/autofeed2.pdf

Also relevant, work on the enumeration of regexps, eg
http://arxiv.org/abs/1204.4982

"a weak measurement extracts such a small amount of information that it leaves the quantum state intact."

That's not correct description -- the quantum state is changed, albeit less than with projective measurement. The paper itself calls it in the abstract "minimal disturbing" measurement, not the "non-disturbing" measurement.

More info here, for example: http://anonymity-in-bitcoin.blogspot.ch/2011/07/bitcoin-is-not-anonymous.html and http://arxiv.org/abs/1107.4524

Bitcoin is not "identity-obvious", but unless extra care is taken, there is a good change that users can be identified and the history of coins traced. An anonymous currency would make that impossible or extremely hard. It seems Bitcoin makes it relatively easy, at least in the current implementation.

He gets 4.5 km/s from his own 2011 paper regarding near-earth asteroids and very specific payload capacities at launch from Earth's surface determining economic returns, which I agree has seemingly arbitrary technology constraints (using Apollo 17 and the MRO Atlas 5.) Somehow that's being distorted into a general statement about asteroid mining, when it's really making specific claims about a single-shot mining operation to and from Earth based on Earth markets, using possibly irrelevant engineering data.

Except you're not taking into account the angle of the plane of the orbit, or where in the orbit the BetaPicB was at the moment the image was captured. ?

From the post you replied to: ...which is separated by about ~ 8 AU at the time..

There's no way of telling.

Yeah, there is, you look up information on the discovery and orbital determination of Beta Pictoris B , and see that it is a near edge on system, with maximum separation occurring around 2013.

s the 25 pixel blob 1.65*Jupiter?

A not so subtle hint is the article saying the blackout disk includes the star, so the, so it should be much closer to the orbit size than the planet size. Nonetheless, ignoring that, and all of the info on the orbit of the planet, if the pixel size was anywhere near the size of Jupiter's diameter, that would be much, much bigger news considering it would mean the telescope is resolving things a factor of a 1000 times smaller than the diffraction limit...

This is ok. For many purposes, software improvements in terms of new algorithms that are faster and use less memory have done more for heavy-dute computation than hardware improvement has. Between 1988 and 2003, linear programmng on a standard benchmark improved by a factor of about 40 million. Out of that improvement, about 40,000 was from improvements in software and only about 1000 in hardware improvements (these numbers are partially not well-defined because there's some interaction between how one optimizes software for hardware and the reverse). See this report http://www.whitehouse.gov/sites/default/files/microsites/ostp/pcast-nitrd-report-2010.pdf. Similar remarks apply to integer factorization and a variety of other important problems.

The other important issue related to this, is that improvements in algorithms provide ever-growing returns because they can actually improve on the asymptotics, whereas any hardware improvement is a single event. And for many practical algorithms, asymptotic improvements are occurring still. Just a few days ago a new algorithm was published that was much more efficient for approximating max cut on undirected graphs. See http://arxiv.org/abs/1304.2338.

If all forms of hardware improvement stopped today, there would still be massive improvement in the next few years on what we can do with computers simply from the algorithms and software improvements.

What happens when Bitcoin goes over $2000? Or what happens when it goes over $50,000? Or what happens when it goes over $200,000? [...] Only time will tell, my friend. Only time will tell.

Good luck with that.

Back in 2008, Adler compared Lageos and LLR estimates of Earth gravity in http://arxiv.org/pdf/0808.0899.pdf and concluded that Earth bound dark matter (outside of the 12,000 km semi major axis of Lageos) is 4 x 10^-9 Earth masses, much below Harris's 5 x 10^-5 estimate (which is based on a comparison with an ancient IAU determination from the 1960's). Given the poor analysis and the discrepancy with the more accurate data used by Adler, I do not think that these results can be given any credence.

I know very little about this field but... Isn't it also possible that our current understanding of the mass of the earth, and its gravity effects, is incorrect and needs to be adjusted?

From New Scientist :

"Harris has yet to account for perturbations to the satellites’ orbits due to relativity, and the gravitational pull of the sun and moon."

That, alone, would make this fail peer review, not to mention that the GPS satellites (which are big and messy and do stationkeeping and get replaced) are not the satellites to use to do this with (the Lageos satellites fit both requirements, being both well monitored and with very low non-gravitational perturbations). The Lageos orbit at a lower altitude than GPS, and so would be expected to exhibit a greater difference with the orbit of the Moon (the other very stable and well characterized Earth orbiter, thanks to Lunar Laser Ranging).

Back in 2008, Adler compared Lageos and LLR estimates of Earth gravity in http://arxiv.org/pdf/0808.0899.pdf and concluded that Earth bound dark matter (outside of the 12,000 km semi major axis of Lageos) is 4 x 10^-9 Earth masses, much below Harris's 5 x 10^-5 estimate (which is based on a comparison with an ancient IAU determination from the 1960's). Given the poor analysis and the discrepancy with the more accurate data used by Adler, I do not think that these results can be given any credence.

More information complexity creates more entropic potential energy.

http://en.wikipedia.org/wiki/Entropic_force
http://arxiv.org/abs/1207.4803
http://arxiv.org/abs/1001.0785

More information complexity creates more entropic potential energy.

http://en.wikipedia.org/wiki/Entropic_force
http://arxiv.org/abs/1207.4803
http://arxiv.org/abs/1001.0785

HR eiter in-company, or out-company (headhunter) loves to play God. Yes they manage payroll, taxes, paperwork, and even a little bit of personalities and do a stellar job. Who doesn't appreciate someone who minimizes potential beef with the IRS. When they play HR they do great, but when they pretend to read minds or futures, they do less well.

Consider engineers, for example. The HR folks are not engineers. They typically are afraid of equations, so algebra, calculus, and differential equations - prerequisites for engineering - are off the table. The only things left are to hope someone else who hasn't used their degree in 15 years (engineering managers) can tell the difference. (It is a good bet that if you haven't used Diff-Eq in 5 years, you aren't any good at it anymore.) HR can try to "hack it" using indirect indicators, GPA, keywords in the resume. As far as I have seen they can be somewhat easily bluffed by a pretty face of a good talker. The problem with that is that it puts folks who are pretty boys with good talk where it should put butt-kicking rock-stars who are going to change the world.

One of my favorite articles, for its lack of implementation or even serious consideration is the ignoble winner: "The Peter Principle revisited". http://arxiv.org/abs/0907.0455

This article gives a really basic, but extensible, framework against things like "frequency of review cycle" can be compared with actual ability to differentiate talent or maximize it in the organization. Use of the structure shows that there are organizational policies and structures that can actually reject talent. You can put in the 100% people, perfect talents, and the performance of the overall organization goes down. If you input some simple parameters you can determine that there is a peak after organizational performance goes down, and never returns to the peak.

This is a sandbox problem that could educate these folks because it gives them some great levers for getting to data-driven understanding.

Yes, or no. Do you deny that given the size of the body of evidence, the probability of ALL available evidence being against the ideas of creationism or "young Earth" is very close to zero? Do you deny that a corollary if this is that SOME evidence must almost certainly be supportive of creationism?

Yes, I deny that any fact supports young earth or old earth creationism. Yes, I deny that there is any evidence supporting young earth or old earth creationism. I'm a denier.

You've repeatedly claimed that some facts support the creationist position. Again, you can either find an example that isn't ridiculously wrong, retract your absurd claim, or keep pulling a "Jane" by doing neither. Since you probably won't surprise us on this account, perhaps lowered expectations are in order. Earlier, you claimed:

Just one example: The fact that radiometric dating relies on certain assumptions has been one of their favorite talking points. Are those assumptions reasonable? I think so. But they ARE assumptions, and that is a fact. Therefore, there do exist facts that can be said to support (or at least not refute) the creationists' arguments. ...

I replied by saying "No, isochron dating only relies on nuclear decay rates being constant, which has been confirmed by SN1987a, etc. Try again." and your response was "Okay, maybe it was a bad example."

Any example may be a bad example, so that wasn't a retraction. Your example actually was a bad example, and anyone who understood my point would have the intellectual integrity to admit that without weasel words. So perhaps my website was down; here's the relevant part:

Isochron dating results of old rocks depend only on nuclear decay rates being constant in time. Isochron dating isn’t dependent on initial quantities of elements, and the analysis method automatically produces error bars on the obtained age. The oldest rocks we have agree that the Earth is 4.55 billion years old, plus or minus 100 million years or so.

Just to be clear, we can’t be sure that nuclear decay rates are exactly constant. But experiments have placed constraints on the size of any variation in decay rates:

1. Supernovae produce many radioactive elements which slowly decay after the explosion. At first they shine brightly in a spectroscopically unique manner, but over the course of several weeks they fade to half their previous brightness. The amount of time it takes the brightness to fade is a direct measurement of the nuclear decay rate. The best example is supernova 1987A, which lies ~169,000 LY away. That means that when scientists looked at that light in 1987, they were measuring the nuclear decay rate as it was around 169,000 years ago. The results were experimentally indistinguishable from current decay rates, and have been confirmed by similar experiments on SN1991T, which is 60,000,000 light years away.
2. The Oklo natural nuclear reactor left evidence that can be used to determine the fine structure constant and neutron capture rates, both intimately entwined with quantum mechanics’ predictions of nuclear decay rates. This experiment is more ambiguous and as a result the error bars are much larger than the SN1987A constraint, but it’s also consistent with a constant nuclear decay rate. Since the Oklo reactor was active 1.8 billion years ago, the Oklo evidence only supports a change in the fine structure constant of one part in 10 mil

Yes, or no. Do you deny that given the size of the body of evidence, the probability of ALL available evidence being against the ideas of creationism or "young Earth" is very close to zero? Do you deny that a corollary if this is that SOME evidence must almost certainly be supportive of creationism?

Yes, I deny that any fact supports young earth or old earth creationism. Yes, I deny that there is any evidence supporting young earth or old earth creationism. I'm a denier.

You've repeatedly claimed that some facts support the creationist position. Again, you can either find an example that isn't ridiculously wrong, retract your absurd claim, or keep pulling a "Jane" by doing neither. Since you probably won't surprise us on this account, perhaps lowered expectations are in order. Earlier, you claimed:

Just one example: The fact that radiometric dating relies on certain assumptions has been one of their favorite talking points. Are those assumptions reasonable? I think so. But they ARE assumptions, and that is a fact. Therefore, there do exist facts that can be said to support (or at least not refute) the creationists' arguments. ...

I replied by saying "No, isochron dating only relies on nuclear decay rates being constant, which has been confirmed by SN1987a, etc. Try again." and your response was "Okay, maybe it was a bad example."

Any example may be a bad example, so that wasn't a retraction. Your example actually was a bad example, and anyone who understood my point would have the intellectual integrity to admit that without weasel words. So perhaps my website was down; here's the relevant part:

Isochron dating results of old rocks depend only on nuclear decay rates being constant in time. Isochron dating isn’t dependent on initial quantities of elements, and the analysis method automatically produces error bars on the obtained age. The oldest rocks we have agree that the Earth is 4.55 billion years old, plus or minus 100 million years or so.

Just to be clear, we can’t be sure that nuclear decay rates are exactly constant. But experiments have placed constraints on the size of any variation in decay rates:

1. Supernovae produce many radioactive elements which slowly decay after the explosion. At first they shine brightly in a spectroscopically unique manner, but over the course of several weeks they fade to half their previous brightness. The amount of time it takes the brightness to fade is a direct measurement of the nuclear decay rate. The best example is supernova 1987A, which lies ~169,000 LY away. That means that when scientists looked at that light in 1987, they were measuring the nuclear decay rate as it was around 169,000 years ago. The results were experimentally indistinguishable from current decay rates, and have been confirmed by similar experiments on SN1991T, which is 60,000,000 light years away.
2. The Oklo natural nuclear reactor left evidence that can be used to determine the fine structure constant and neutron capture rates, both intimately entwined with quantum mechanics’ predictions of nuclear decay rates. This experiment is more ambiguous and as a result the error bars are much larger than the SN1987A constraint, but it’s also consistent with a constant nuclear decay rate. Since the Oklo reactor was active 1.8 billion years ago, the Oklo evidence only supports a change in the fine structure constant of one part in 10 mil

Ha, burn!

The original physics were explained in 1967, per the footnote of the article.

http://arxiv.org/ftp/arxiv/papers/1312/1312.4006.pdf

This depends on the field. In physics and astronomy, almost everything gets published on arxiv.org (in addition to the journals), where it's free to access.

Well, it's sort of complicated.

On one hand—it's probably worth pointing out that the American punitive system is absolutely insane, and the mildness of this should not be taken as evidence of a defective process simply because it doesn't follow suit. Indeed, there are some fairly involved legal and philosophical reasons as to why the punishments aren't more extreme. Here is a paper on it. (I haven't read all of it, but it seems sensible enough from the first few pages.) One of the key points is that a lot of money goes down the toilet on dead ends and genuine errors anyway; another is that scientific misconduct isn't actually illegal, so the power of funding bodies to defend themselves is somewhat limited. In the end, the top priority is still getting them out of science.

The thing you need to be careful of is mixing Newtonian ideas (instantaneous force vectors) with quantum or relativistic phenomena. However, there are many similarities between gravity and electromagnetism. If you go through all the math and consider the propagation delay between a moving charged particle and one at rest, it works out that the force vector points towards the instantaneous position of the moving charged particle because the EM field of the moving particle has a velocity-dependent component that cancels out the propagation delay effect. A similar thing happens if you work through the uglier equations of General Relativity, but the cancellation effect is not exact, and this difference is what gives rise to gravitational radiation (and is why we believe orbital pulsar systems have decaying orbits), and is also why we do not seem to detect a gravitational aberration effect..

This paper explains it pretty well. The math isn't too bad, though it uses tensor notation so it can look a bit intimidating if you're not used to it, but the text around the equations is pretty good.

Agreed. I'll answer your question in a moment ...

I've been online for 20+ years. The cycle of online human communication is *always* the same.

* Forum (BBS / newgroup / website) provides a common ground for people to share info. / tips / opinions
* Site is small as only the "geeks" use it
* Site gains Momentum and goes mainstream
* The crazies come out of the woodwork -- Name Calling / Ad Hominem / Trolls attack -- people keep forgetting authority needs accountability
* Moderators are either
    a) 45% of the time non-existent
    b) 45% of the time grammar/spelling/free-thought nazis where posters need to tow the party line,
    c) 4% of the time does it rarely exist there are balanced moderators who allow a difference of opinion as long as it it kept civil and intelligent
    d) 1% site allows members to self-moderate
* All the old members complain about "the good ole days" when the noobs / newbs / hipsters, etc. didn't drive the S/N from Signal into Noise
* New site starts that promises to be "Bigger, Better, Cheaper", etc.
* Old site membership is split as some members leave to check out "Awesome new site" (temporarily, others for good),
* Old site lingers but never really recovers from the mass influx of growth and decay.
* Rinse and Repeat ad nauseum.

What /. did innovate at the time was to allow the crapfest of usenet to be FILTERED. Reddit has mob rule when you get carpet modded into oblivion because people don't want their thinking challenged.

Newspaper used to exist because people saw the value in someone else filtering the amount of information to collect mostly signal and to present THAT to you so you didn't have to waste your time filtering the S/N.

Now to answer your question:

Why is *good* moderation so HARD?
  1. Because it involves TRUST. Are you an expert? Prove it? etc.
  2. The problem is that Truth is NOT only objective, but ALSO subjective. The majority fall into the fallacy of duality. "I'm right, THEREFORE you're wrong." instead of being humble and honest enough to admit. "My POV has + and -, Your POV has + and -. What *new* things can we learn from the difference and intersection of these strengths and weaknesses?"

As a Mystic I am able to see the Strengths and Weakness in *everything*. The question is NOT about simple-minded good vs evil, but about being able to have an open mind and consider ALL the possibilities: the short-term, the long-term, how the strengths of short-term thinking/action might eventually become the negative in the long-term, and vice versa, what did the negative teach us, etc. Most people are not able to communicate with clear, simple, logic free of mis-guided emotion, let analyze something to that depth.

Being passionate is fine. Be able to walk the line between Logic and Emotion -- yeah, we're all still trying to figure that one out. Especially when some noob / fanboi makes an ignorant comment and you just want to flame his ass for being a stupid git. :-)

cue oblg. xkcd ...
  http://xkcd.com/591/
http://3d.xkcd.com/802/

References:
  * "A Community Membership Life Cycle Model" http://arxiv.org/pdf/1006.4271.pdf
  * http://en.wikipedia.org/wiki/Online_community

Earth formed from heavy elements produced in at least one prior generation star, but there could be more than just one prior generation (It's very, very unlikely that all or nearly all the medium weight elements above lithium, in our solar system, came from just one older star, and pretty unlikely even that all the heavies above Iron were cooked up in just one supernova).
          It's not a safe assumption that stars last an average of 10 Billion years. The most numerous types, red dwarfs, make up 80-90% of all stars, last a lot longer than that, and probably stay stable on the main sequence for 100-200 billion years (American Billions). They also shouldn't spread elements around much when they finally do leave the main sequence. Stars about the size of our Sun, spectral class G2, typically live about 10 Billion years, but make up only about 2% of stars. Big stars, type O, B, and A, burn more quickly, and it's possible to get enough hydrogen together for a star to burn through all its fuel and supernova in mere hundreds of thousands of years, or possibly even a blazing fast 10's of thousands. Those stars are rare, but they are so massive that even a few produce enough heavy elements and push enough gas around when they supernova, to create hundreds of sun sized and smaller stars and all the heavy elements to give such stars the solid, rocky planets we now think are practically ubiquitous.
            The supernova explosions are a common source for two effects - heavy element formation, and compressive shock waves that trigger new stars forming in nearby interstellar gas clouds. Many of these gas clouds are already enriched with heavy metals from previous supernovae, Spiral galaxies tend to get regions of new star formation, and quiet regions. But, the high and low density regions in spirals like our Milky Way exist on larger scales than the star forming "nursery" clouds, and this is largely because gas clouds are not just compressed by novae - both the dense star forming clouds and very large but more difuse clouds colide with other clouds, including clouds that were part of dwarf galaxies being captured by the big spirals. So, it's a partial coincidence - Older generation stars have some influence on the shapes of spiral galaxy features, but dwarf galaxy capture has more, and the rare colisions of spirals with other big galaxies show just how much influence the large scale objects can have, producing wildly twisted galaxys such as

          If anyone wants to read up on this sort of thing, please remember, because astronomers named them before they knew anything about why there were multiple distinct types of stars in the same mass ranges, Population II stars are actually older than Population I, and Population III older than II. A given population usually includes multiple generations of stars. As an exception, the very oldest, massive stars that novaed within the first million years or so after star formation began, and produced so many heavy elements are called Population III, and most probably represent just a single generation and possibly only the largest types.

https://en.wikipedia.org/wiki/Stellar_nursery#Stellar_nurseries

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

https://en.wikipedia.org/wiki/Population_I_stars#Population_I_stars

and, for those people wanting more than just the Wikipedia versions, a little real source material:

http://arxiv.org/pdf/astro-ph/0012399v2.pdf

As the summary notes, it was only recently discovered that this system has three gravitationally bound stars, making it the widest such group currently known. Paper from a few months ago on the arXiv, and a news write-up of that.

Yep, except it's total bullshit.

Apologies, forgot to actually hyperlink...

2nd sentence in the introduction is wrong. Makes it hard for me to want to finish reading the rest of it.

Apologies, forgot to actually hyperlink...

The article is free on arXiv like a vast number of physics and astronomy articles.

Penrose is one of the most prominent respected scientists to propose noncomputable physics (see his "The Emperor's New Mind" and "Shadows of the Mind"), but his arguments were refuted by various experts, from physicists to logicians (see, for example, "A Refutation of Penrose's Godelian Case Against Artificial Intelligence"). He seemed to do it without any evidence, really, and claimed that quantum mechanics must be wrong since it's computable (as is, by the way, quantum field theory).

There were other related arguments around the same time bearing upon this. One had to do with things such as a proposal that artificial neural networks can be super-turing if the weights at nodes could be arbitrary precision real numbers--something clearly excluded by the Bekenstein bound. Yet another one was the observation that a turing machine modified to have continued interaction with its environment, rather than just starting with input and ending with output, can be super-turing. However, this argument is akin to a slight of hand, because it's based on a misleading definition of what the information processing system is, by excluding part of it--the interacting environment in question. If that environment is finite (and it is limited by the system's light cone, unless you believe superluminal communication speed is possible), then the combined system is still sub-turing: a non-deterministic LBA.

So are human minds subject to the same theoretical limits that apply to finite computers? It certainly appears that way, according to our current understanding of physics. Of course, it may be that this will be upturned in the future by the discovery of new, noncomputational physics, and, of course, not all philosophical worldviews align with physicalism.

As for your comment on how powerful quantum computer algorithms can be, we already have a pretty good idea that they're not all that powerful--they're just more efficient (and, specifically, polynomially more efficient); see the computational complexity classes defined for them, like BQP etc. http://arxiv.org/abs/0804.3401

http://arxiv.org/ftp/arxiv/papers/1304/1304.3381.pdf Extrapolating life origin back to early universe

Storage / RAM is not the only application for memristors. As they can serve as single-element excitation counters, they enable extremely power-efficient neuromorphic chips, as described in this paper from Intel (PDF warning): Proposal For Neuromorphic Hardware Using Spin Devices

We present a design-scheme for ultra-low power neuromorphic hardware using emerging spin-devices. We propose device models for 'neuron', based on lateral spin valves and domain wall magnets that can operate at ultra-low terminal voltage of ~20 mV, resulting in small computation energy. Magnetic tunnel junctions are employed for interfacing the spin-neurons with charge-based devices like CMOS, for large-scale networks. Device-circuit co-simulation-framework is used for simulating such hybrid designs, in order to evaluate system-level performance. We present the design of different classes of neuromorphic architectures using the proposed scheme that can be suitable for different applications like, analog-data-sensing, data-conversion, cognitive-computing, associative memory, programmable-logic and analog and digital signal processing. We show that the spin-based neuromorphic designs can achieve 15X-300X lower computation energy for these applications; as compared to state of art CMOS designs.

I research hard AI. In my view thinking through and tackling example problems is the best way to explore a topic. If you require your system to mirror our current understanding of neuroscience, then you're essentially researching the algorithms of the brain.

If you're specifically looking into epilepsy and related, consider checking out William Calvin's website. He's an experimental neuroscientist from University of Washington, who wrote many books that explain the neurological foundations of the brain in readable form with good detail.

(1) What are some interesting computational neuroscience simulation problems

Pretty much anything AI falls under that category. Go over to Kaggle.com and check out some of their competitions, including their past competitions. Check out the Google AI lab and see what they're doing, and check out recent publications to see what people are trying to solve. Ask yourself: Are humans better than the computer, and can it be done better?

Here's a video of a system that uses neuron simulation (of a sort) to recognize hand-written digits. A hand-written digits dataset is in the UCI archive below.

(2) Is it easy for a non-academic to get the required data?

Generally, yes. UCI has a repository of machine-learning datasets. The researchers supporting Kaggle competitions frequently release their data.

I've found that researchers are generally approachable, and will give away copies of their data (I have 4 datasets from researchers). As a personal anecdote, last week a researcher from this very forum sent me his dataset of Mars altitude images - I'm trying to come up with an algorithm to recognize craters.

(3) I am familiar with (but not used extensively) simulators like Neuron, Genesis etc. Other than these and Matlab, what other software should I get?

In my view, pick a computer language that has a wide support network of libraries, and code things from scratch.Something like Perl or R. At some point you will want to break open the box and see what's actually happening inside, and familiarity with the system (having constructed it) is key. You will want to insert trace statements, print out intermediate results, and so on. Most of the pre-built systems don't have what you will ultimately want, and building simulation objects isn't terribly hard.

(4) Where online or offline, can I network with other DIY Computational Neuroscience enthusiasts?

Please let me know if you find any (by posting a response).

I've found that most AI enthusiasts are really "big data" enthusiasts, and most of them are all about business rather than AI. The IRC AI chatrooms are all but dead, and most of what is there are students asking for help with their homework. (Although to be fair, the lurkers there know everything about AI and can answer questions and make suggestions if you're stuck.)

The NEAI meetup in Cambridge is mostly spectators - people who want to find out about AI or how to use AI ("how can I use AI to improve the performance of my financial company?"). I hear there's an AI meetup out on the West coast that's pretty good.

See if there's a meetup in your area for something related, or start one and see if anyone shows up.

The /. summary is straight from the blog, and really, the only interesting part of the blog post.

Here's the actual paper (PDF): http://arxiv.org/ftp/arxiv/papers/1311/1311.2526.pdf

eLisa (a toned-down, less costly, version of the original LISA) requires LISA Pathfinder (a technology demonstration mission). It turns out that LISA Pathfinder can be used to perform the first solar system test of MOND. This subsidiary test, largely flying under the radar, could potentially confirm (or rule out) the MOND / TeVeS hypothesis, providing an important test of fundamental physics for remarkably little expense.

eLisa (a toned-down, less costly, version of the original LISA) requires LISA Pathfinder (a technology demonstration mission). It turns out that LISA Pathfinder can be used to perform the first solar system test of MOND. This subsidiary test, largely flying under the radar, could potentially confirm (or rule out) the MOND / TeVeS hypothesis, providing an important test of fundamental physics for remarkably little expense.

For those interested in a more in-depth treatment... http://arxiv.org/pdf/1306.3008.pdf

http://arxiv.org/abs/1306.3008 http://arxiv.org/pdf/1306.3008v2

http://arxiv.org/abs/1306.3008 http://arxiv.org/pdf/1306.3008v2

Arxiv to the rescue: http://arxiv.org/abs/1306.3008 (This may lack editorial changes etc. made by the journal, but should be factually complete.)

Of course the "academics" will be the last ones to abandon their bogus and wrong practice of artificial scarcity of information.

Yeah, academics certainly want nobody to know their work ... sure. That must also be the reason why so many scientists put up their published articles (and often even preprints which are not yet accepted in a journal) on places like arXiv.

Making information available is just one function of journals, and not even the most important. The most important service journals provide is peer review. So if you want to replace journals, don't think about how to make articles available. That problem has already working solutions. Think about how to either organize peer review differently, or how to better organize peer review, or how to replace it with something which is better. Note that one of the important points is that on one hand, the reviewers should have the expertise needed, and ideally people with known interests that the article does or does not appear are excluded. Therefore you cannot allow just anyone to review the article; indeed, much of what goes into the reputation of a journal is how well they choose their reviewers. On the other hand, the author should not know who reviewed the paper, to allow the reviewer to be honest even if it is a paper by an influential author. As reviewer, you have to be able to say that an influential author has written shit without fear that you risk your career for it.

Also note that currently Open Access journals require payment from the author to get the article published, and the fees are not that low. Which has its own share of problems:
* Since now the income of the journal doesn't any longer depend on the number of readers (and thus on the quality of the articles) but on the number of published articles, there's the danger that an open access journal lowers its quality standards in order to get more articles published and thus paid for.
* While OA journals don't restrict reader access, the publication fee creates a barrier for authors who only can publish if they can pay the money (either by being part of an institution which pays it for you, or by having enough money yourself to be able to pay it). I'm not sure that Einstein would have been able to pay for a publication in an Open Access journal from his patent officer's fee.

Yet, somehow people think selling 1's and 0's to people with computers is a valid business strategy -- And we let them call themselves "scientists".

You just showed that you have not the slightest clue about how academic publishing works. The scientists get not a single cent from the publisher for their articles. Let me repeat that, so that this is absolutely clear: The researcher does not get paid by the journal for his article. The scientist profits only and exclusively through the reputation the publication gives him (and the indirect effects, like being able to get funding).