Header 1

Our future, our universe, and other weighty topics


Thursday, August 17, 2017

AI Alarmism Is Unwise Because Understanding Is Not a Digital Output

Computer experts have long struggled to achieve artificial intelligence (AI), computers or robots with human-like intelligence. There have been many errors in predictions about when artificial intelligence would appear. Page 12 of the paper here gives a graph showing that 8 experts predicted that computers would have human level intelligence by about the year 2000, and that 8 other experts predicted that computers would have human level intelligence by the year 2020 (something incredibly unlikely to happen in the next few years).

Some experts continue to make dubious predictions about artificial intelligence. Some have become what we may call AI alarmists. An AI alarmist is someone who warns us that computers or robots are going to get so smart that a great disaster will occur. Some say that a large fraction of the population will become out of work, as computers or robots take the jobs of bankers, physicians, lawyers, and software developers. Other AI alarmists predict something worse: that machines will become so smart that they take over the planet.

Such alarmists often say, “Maybe if we're lucky they'll keep us as pets.” High-tech luminary Elon Musk has said, “A.I. is a fundamental risk to the existence of human civilization, and I don’t think people fully appreciate that.” Stephen Hawking has made similar comments.

According to the computationalism theory of the human mind, the mind is like a computer, so one day we will be able to develop computers that produce outputs just like human consciousness. Such a theory is assumed by most AI alarmists. Such theorists usually don't tell us that they are advancing the computationalism theory of the human mind. They usually just pronounce the dubious ideas of such a theory as if such ideas were self-evident.

But the computationalism theory of the human mind is not valid. The human mind is not like a computer, and the brain has nothing like these seven things that a computer uses to store and retrieve data. The human mind has facets such as conscious experience and understanding, which have never been produced to any degree by a computer.

Let us look into what happens when computers compute. The following equation covers most of the types of computation that occur.

digital inputs + processing = digital outputs or modification of digital data

There are various types of variations of this equation. One is simply:

no inputs + processing = digital outputs or modification of digital data

Another variation is the following:

digital inputs + processing + retrieval of other digital inputs = digital outputs or modification of digital data

By digital inputs or digital outputs I mean anything at all that can be represented digitally, by a sequence of binary numbers. Here are some of the things that we know can be represented digitally, and which modern computers do use as digital inputs or digital outputs:

Any number
Any set of characters or words
Images
Videos
Databases

Any text can be digitally represented by means of things such as the ASCII system that allows you to represent particular characters as particular numbers. While we don't normally think of an image as digital, it can be represented digitally as a series of pixels or picture elements. For example, a photograph might consist of 1 million pixels, which each can be represented by a number representing a particular shade of color. So the image can be digitally represented by a million such numbers. A video can also be representing digitally, since the video can be represented as a series of images, each of which can be digitally represented.

But there are some things that we can never hope to produce as digital outputs. The first is real conceptual understanding. By understanding I don't mean “how-to” type understanding, but the high-level conscious understanding of some abstract truth or concept. We can imagine no possible way to produce a digital output that would equal a real conceptual understanding of something.

But, you may ask, doesn't that smart computer Watson already understand something – the game of chess? No, it doesn't. Watson merely can produce a digital output corresponding to a good move to make as the next move in a chess game. Watson has zero conceptual understanding of the game of chess itself, and has zero understanding of the abstract concept of a game. The only way you can understand the abstract concept of a game (or the abstract concept of leisure) is if you have been a human being (or something like a human), and played a game yourself.

A digital output must always boil down to a series of 1's and 0's. Can we imagine a series of 1's and 0's that would equal a real understanding of an abstract concept such as health, matter, life or world peace? No, we cannot.

AI alarmism is based on the idea that future computers will be able to produce conceptual understanding as an output. They won't, because real understanding of abstract concepts is not a possible digital output, and digital computers will only be able to produce digital outputs. Computers or robots lacking conceptual understanding will neither be able to take over the world nor even ably perform any of the more intellectually demanding jobs requiring the repeated application of general intelligence.

But why do computers sometimes seem smart? Because by programming software and putting that into a computer, a computer can act as a repository of human logic. But the logic used by the computer is not coming from the computer, but from some human who programmed the computer. The process of encoding human logic and transferring human logic to a computer is relatively slow and laborious, only allowing the simulation of very limited types of expertise. There seems to be no hope that clever humans will ever be able to create some kind of general intelligence program that thinks and analyzes in the general-purpose way that humans do.


Faced with such reasoning, an AI alarmist may reply with clever reasoning like this:

But we know that the brain produces understanding, and the brain is a material thing. Once we understand the exact material factors involved in how the brain produces understanding, we need merely ramp up such physical processes in a robot or a computer, multiplying such processes many-fold; and then you'll have something that greatly exceeds us in intelligence and understanding.

I deny that we know any such thing as what is stated in the first sentence of this argument. Nature never told us that our thoughts and ideas are coming from our brains. The idea that the understanding of the human mind is produced by the brain is an unproven dogma – something very often asserted, but never proven. Such a dogma is certainly not proven by brain imaging studies, which merely show very unimpressive differences in blood flow to different parts of the brain, typically only 1 or 2 percent (see here for the flaws of brain imaging studies).

Below are eight reasons for doubting that human understanding is merely a product of the brain.
  1. No one has any real understanding of how neurons or any other brain parts could produce consciousness, ideas or understanding.
  2. As documented by physician John Lorber and others, and as discussed here, there are numerous cases in the medical literature of people who maintained normal or almost normal consciousness and understanding, even though very large parts of their brains were destroyed or ruined by disease or injury.
  3. The mammal dissection experiments carried out over many years by Karl Lashley showed surprisingly high mental functionality when large portions of animal brains were removed, including large fractions of the cortex.
  4.  As argued here, the human mind has quite a few fundamental traits that we cannot explain as being caused by natural selection, because they don't provide survival value; and this undermines the prospects of explaining our minds as some material effect.  
  5. The human mind occasionally displays psychic powers (as demonstrated here and here) that cannot be accounted for through brain activity.  
  6. During near-death-experiences there have been numerous cases of people who reported floating out of their bodies during cardiac arrest, and often correctly reported details during a time when they should have been unconscious.
  7. Claims that understanding comes from the brain (or more specifically the brain cortex) are in conflict with tests showing very high mental functioning and apparent high understanding in animals such as crows, who have no cortex and tiny brains.
  8. Human memory is still supremely mysterious, and we have no understanding of how brains could be achieving the 50-year memories that humans demonstrate, or the instantaneous recall of memories that humans have. Despite the dogma that memories are stored in brains, there is no plausible neural explanation as to how an organic system like the brain could be the source of memory storage as long-lasting as humans have, or the source of memory recall as instantaneous as humans have.
All of these things suggest an idea much more logical than the “we are thinking meat” dogma of materialists: the idea that we are thinking souls who happen to be hanging around in bodies. Because this idea is highly viable, we do not at all know that understanding is something that is materially produced by human brains. Not knowing such a thing, we can have no confidence that some “trick of matter producing mind” will ever be uncovered by future scientists.

If you are “thinking meat,” then you might have a grave fear that maybe the secret of a “meat-to-mind” trick might be learned by computer makers, who might be able to amp up such a trick a thousand-fold, to produce robots and computers so smart that they take over all jobs such as yours, or take over the world. But in light of the eight things discussed above, a more logical idea is: there is no such “meat-to-mind” trick for us to ever discover, because the meat in our brains is not producing our minds. In that case, we need not fear very much robots or computers, because they simply will never be able to have any bit of the understanding we have. Computers or robots lacking any real understanding will not be able to perform any of the more intellectually demanding jobs, those that require real conceptual understanding. And such robots and computers will not be able to take over the world, having no actual understanding of our planet or even a thousand simpler things.

Sunday, August 13, 2017

Shocking CMB Anomalies Contradict Guth's “Empirical Success” Claims

Earlier this year, Scientific American published a blistering critique of the theory of cosmic inflation originally advanced by Alan Guth (not to be confused with the more general Big Bang theory). The theory of cosmic inflation (which arose in 1980) is a kind of baroque add-on to the Big Bang theory that arose decades earlier. The Big Bang theory asserts the very general idea that the universe began suddenly in a state of incredible density, perhaps the infinite density called a singularity; and that the universe has been expanding ever since. The cosmic inflation theory makes a claim about less than one second of this expansion – that during only the first second of the expansion, there was a special super-fast type of expansion called exponential expansion.

The article in Scientific American criticizing the theory of cosmic inflation was by three scientists (Anna Ijjas, Paul J. Steinhardt, Abraham Loeb), one a Harvard professor and another a Princeton professor. It was filled with very good points that should be read by anyone curious about the claims of the cosmic inflation theory. But now the article has been half-censored, for Scientific American has put the article behind a paywall. But don't worry, you can still read the article on a Harvard web site here. Or you can go to this site by the article's authors, summarizing their critique of the cosmic inflation theory.

The Scientific American article by the three scientists provoked an unusual response. The main supporters of cosmic inflation theory (including Alan Guth and Andrei Linde) along with about 30 other cosmologists have published a rebuttal article called “A Cosmic Controversy.” It is kind of an authoritarian power play, designed to impress the reader by listing authorship by some of the top names in cosmology. The list of authors is very impressive, but there are some factually inaccurate claims in the article.

The article claims that the cosmic inflation theory has been empirically successful. Referring to variations of the theory called slow-roll models, Guth claims that “many models in this class continue to be very successful empirically,” and later refers to “the dramatic observational successes of inflation.” These claims are not accurate.

The cosmic inflation theory makes two main claims: (1) that there is or was something called an inflaton field; (2) that the universe underwent a period of exponential expansion during part of its first second. Various versions of the cosmic inflation theory have also claimed that there exist or once existed other universes (sometimes called “bubble universes”), or that something called grand unification theories (GUT theories) are in some sense correct. We have not a single observation verifying any of these claims.

The claim that the universe underwent an exponential expansion during part of its first second is one that can never be confirmed, because of physical limitations (the recombination era issue) that will always prevent us from looking back to the universe’s first second with our telescopes. No evidence has been produced for this so-called inflaton field, and it has not been detected by the Large Hadron Collider. No evidence has been produced that there is any universe other than our own, and there are very strong reasons for thinking that no such evidence ever could be produced (anything that we might observe would always be an observation of our universe, not some other universe). No evidence has turned up for the idea of grand unification theories (GUTs), which have been one of the more embarrassing failures of modern physics.

Given this situation, in which not a single one of its key claims has been backed up by observations, it is pure baloney for Guth and his allies to be calling their theory an empirical success. Their claim is based on the very shaky idea that various versions of cosmic inflation theory have made some predictions about a few things that have been consistent with observed reality. In my post here I give several paragraphs explaining why this type of claiming success based on imprecise general predictions is fallacious. Among the reasons I gave is that there are very many versions of the cosmic inflation theory (each predicting a wide range of things, as each version allows a range of input parameters); so even if the theory is bunk we should not be surprised if some of the resulting predictions matched reality.

Let us suppose that some theory claimed that green monsters from Alpha Centauri have landed on our planet and invested in the bond market; consequently next year the bond market will go up and the stock market will go down. If next year the bond market did go up and the stock market did go down, it would still not at all be accurate to claim that this weird theory was empirically successful. It would only be correct to say that if the green monsters were actually observed. Similarly, no theory of cosmic inflation can be called empirically successful until one of its core central claims (unique to the theory) has been observed. No such thing has happened.

theory predictions

There is another huge reason why the cosmic inflation theory cannot be called empirically successful. The reason is that the theory is inconsistent with observations of anomalies in the cosmic background radiation (also called the cosmic microwave background, or CMB). 

Believed to date from early in the universe, the cosmic background radiation is a type of radiation pervading all of space. The cosmic inflation theory predicts that this radiation should be very smooth and isotropic (just as the theory that your friend spent 20 minutes stirring his bowl of pancake batter predicts that the pancake batter should be very smooth, and without lumps).

The first satellite to observe in detail the cosmic background radiation was the WMAP satellite launched in 2001. This satellite detected some very strange anomalies in the cosmic background radiation, anomalies that came as a surprise to scientists. One was an anomaly called the cosmic cold spot. Another was an anomaly that is technically known as the hemispherical variance asymmetry. Then there is an anomaly called the quadrupole-octopole alignment. There are nine other anomalies in the cosmic background radiation that are summarized in a table in this scientific paper. The table is below:


The p-values here give us a rough idea of the probability of finding such anomalies if standard ideas of cosmology (including cosmic inflation and dark matter) are correct. This all presents a huge problem for cosmic inflation theorists such as Guth. These are all things that we should not expect to be finding in the cosmic background radiation if the cosmic inflation theory is correct.

Years after the WMAP satellite was launched, scientists launched another satellite called the Planck satellite. It was predicted that the troubling anomalies in the cosmic background radiation would go away after the more powerful Planck satellite did its work. But that did not happen. The Planck team reported the same anomalies. The table above is from a paper entitled, “CMB Anomalies After Planck.” When the Planck team reported its results on these anomalies, they buried their findings in a dense technical document, as if they were trying to make it as hard as possible for anyone to discover the truth about this matter. But the “CMB Anomalies After Planck” paper gives us some of the straight talk that the Planck paper lacked.

Referring to the hemispherical asymmetry it reported, the paper says, “An inflationary theory could, in principle, accommodate models that produce hemispherical asymmetry, but such a model would have to be multi-field and involve, for example, a large-amplitude superhorizon perturbation to the curvaton field.” Which is a fancy say of saying the simpler versions of the cosmic inflation theory are not compatible with this CMB anomaly – only more baroque and implausible versions are compatible with it. Requiring a multi-field cosmic inflation theory (imagining not just one but more than one undiscovered fundamental field involved in primordial cosmic inflation) is kind of like it being that your theory of crop circles requires not just UFO's but also cooperation by Bigfoot creatures.  In a similar vein, this paper states that the most popular version of the cosmic inflation theory (called slow-roll inflation) is not compatible with the CMB anomalies:

Measurements of CMB temperature fluctuations by the Wilkinson Microwave Anisotropy Probe (WMAP) indicate that the fluctuation amplitude in one half of the sky differs from the amplitude in the other half. We show that such an asymmetry cannot be generated during single-field slow-roll inflation without violating constraints to the homogeneity of the Universe.

But this “slow-roll” version of the inflation theory is the very one that Guth called “very successful empirically” in his Scientific American article. Apparently it is no such thing.

A 2010 paper states the following about the anomalies in the cosmic background radiation:

While not all of these alignments are statistically independent, their combined statistical significance is certainly greater than their individual significances. For example, given their mutual alignments, the conditional probability of the four normals lying so close to the ecliptic, is less than 2%; the combined probability of the four normals being both so aligned with each other and so close to the ecliptic is less than 0.4% × 2% = 0.008%. These are therefore clearly surprising, highly statistically significant anomalies — unexpected in the standard inflationary theory and the accepted cosmological model.

This is a probability of less than 1 in 10,000 under the assumptions of the cosmic inflation theory and the accepted cosmological model. That's hardly what we would find if the cosmic inflation theory really was “empirically successful” as Guth claims.

Using the phrase “in tension” to mean “conflict with,” Stephon H. Alexander (a professor of physics at Brown University) writes the following about these anomalies in the cosmic background radiation, and their relation to the cosmic inflation theory:

During the epoch that the CMB anisotropies were formed, they too are supposed to, on average look the same in every direction. The theory of cosmic inflation generically predicts this feature. This means that if one divides the sky into two arbitrary hemispheres, we should see the same statistical features of the anisotropies in both hemispheres. However, both WMAP and Planck see a difference in the amount of anisotropies in different hemispheres in the sky. This feature is in tension [with] one of the most powerful attributes of inflation, whose rapid expansion of space-time smooths out any large-scale directional preference, while democratically sprinkling the space-time fabric with the same amount of ripples in every direction. With some decorative tweaking, it is possible to modify inflation to account for the anomaly, but this seems to be at odds with what inflation was invented for-to make the early universe smooth enough and see the tiny anisotropies that later become galaxies. One might think that this would be a great opportunity for alternative theories of the early universe, such as bouncing/cyclic cosmologies to rise to the occasion and explain the anomalies, but so far, there is no compelling alternative.

By suggesting that it may be time for “alternative theories of the early universe,” Alexander is clearly suggesting these CMB anomalies are in conflict with the cosmic inflation theory.

Contrary to the claims of Guth and his clique, the cosmic inflation theory is not empirically successful. It has enjoyed another type of success: sociological success. The history of modern science culture shows repeatedly that a theory that is not empirically successful may become sociologically successful and become popular due to a bandwagon effect and groupthink. Once this snowball effect gets rolling, the theory may become a speech custom of an insular academic subculture, and a little piece of tribal folklore has been born. The adherents of the theory will in effect place gold medals around their own necks, congratulating themselves on what they think is their brilliant explanatory triumph. But such gold medals may be very undeserved.

But why did I use the term “shocking” in this blog post's title to refer to these anomalies in the cosmic background radiation? There's one CMB anomaly that is quite shocking. It seems that something called the quadrupole – octopole alignment aligns with the plane of our solar system. Since the cosmic background radiation has often been described as something that looks the same everywhere in the universe, we should not expect to find any such correlation involving our solar system. One paper states the following:

Particularly puzzling are the alignments with solar system features. CMB anisotropy should clearly not be correlated with our local habitat. While the observed correlations seem to hint that there is contamination by a foreground or perhaps by the scanning strategy of the telescope, closer inspection reveals that there is no obvious way to explain the observed correlations. 

Physicist Lawrence Krauss has this to say about this topic:

 But when you look at CMB map, you also see that the structure that is observed, is in fact, in a weird way, correlated with the plane of the earth around the sun. Is this Copernicus coming back to haunt us? That's crazy. We're looking out at the whole universe. There's no way there should be a correlation of structure with our motion of the earth around the sun — the plane of the earth around the sun — the ecliptic. That would say we are truly the center of the universe. The new results are either telling us that all of science is wrong and we're the center of the universe, or maybe the data is simply incorrect, or maybe it's telling us there's something weird about the microwave background results and that maybe, maybe there's something wrong with our theories on the larger scales. 

There is a small band of cosmology-following geocentrists who believe that the Earth is the center of the universe, and that this quadrupole – octopole alignment supports their claim. But we know the earth revolves around the sun. A less outrageous claim would be that the solar system may be somehow in some kind of privileged position, and that the quadrupole – octopole alignment supports this claim. It would be premature to make even this less outrageous claim based on this limited evidence. But it is interesting that this quadrupole – octopole alignment may suggest one of the key assumptions of modern cosmology (the Copernican principle, that there is nothing special about the position of the solar system) may be wrong. 

Postscript: For more information on this topic, do a Google search for "lopsided universe," "axis of evil (cosmology)", "CMB cold spot," and "CMB anomalies." In a previous post I  had mentioned the cold spot as a thorn in the side of cosmic inflation theory, but from the table above it is clear that the hemispherical variance asymmetry is a ten times bigger problem for that theory. 

Wednesday, August 9, 2017

No, Kurt, the Hippies Didn't Plant the Seeds of Trump's Election

One of the standard plays of those who refuse to accept observations of the paranormal involves assuming a sanctimonious “holier-than-thou” attitude in regard to rationality. The skeptic will imply that he and his ideological allies are rational thinkers, and that those who believe otherwise are irrational. A lame example of this strategy is to be found in a long essay recently published in The Atlantic. The essay by Kurt Andersen is called “How America Lost Its Mind.” The approach Andersen takes is to go throughout the past 60 years of American history, and belittle a huge variety of observations, theories or thought tendencies that he dislikes. He tries to trash all such things by claiming that they were irrational, part of a process of America losing its mind. The resulting critique, involving a great deal of unfairness and poor logic, is a kind of toxic soup that is served up in a cup marked “100% rational.”

Andersen disparages the counterculture of the 1960's, trying to portray it as some outburst of irrationality and reality-denial. This is quite unhistorical. The hippie-movement or counterculture of the 1960's largely arose as a reaction to the sins of 1960's America, which included racial discrimination, conformist consumerism, and the waging of a senseless war abroad in Vietnam in which more than 300,000 died or were maimed by bombs dropped by American bombers and toxic Agent Orange sprayed by American personnel. Reacting against such things was more like rationality and paying attention to reality than the opposite.

Two of the visuals of Andersen's article show peace signs and an “End the War” sign among various other portrayals supposed to show irrational Americans. Andersen says, “As the Vietnam War escalated and careened, antirationalism flowered,” and then in the next sentence he discusses a Vietnam war protest described by Mailer. The net result is we are left with the insinuation that opposing one of America's most senseless wars was an act of irrationality. But it was actually the opposite of that. 

 1960's: a rational response to an irrational war

Andersen then attempts to make a list of various things that he thinks are examples of America “losing its mind” and becoming “untethered from reality.” Among the things he lists are Raymond Moody's book describing near-death experiences, and Charles Tart's research on out-of-body experiences. But what are these things doing in such a list? Both near-death experiences and out-of-body experiences are occasional parts of human experience, whatever their cause. Documenting such experiences is an example of paying attention to reality, not becoming untethered from it. Andersen gives us a bogus claim that Charles Tart “proceeded to devote his academic career to proving that attempts at objectivity are a sham and magic is real.” To the contrary, Tart is a very serious scientific researcher who has written many level-headed books dealing with parapsychology research. I've read several of his books, and they never presented any belief about magic. You can see Tart's recent blog postings here, and you will find no sign of Andersen's cheap-shot depiction being correct.

Andersen then goes on to include UFO observers in his mudslinging, by saying, “I’m pretty certain that the unprecedented surge of UFO reports in the ’70s was not evidence of extraterrestrials’ increasing presence but a symptom of Americans’ credulity and magical thinking suddenly unloosed.” But there's no “magical thinking” in a typical UFO sighting – a person simply reports some strange thing he saw. Absurdly, Andersen tries to suggest that maybe we should have denied Jimmy Carter the presidency because he saw a UFO. He states, “Until we’d passed through the ’60s and half of the ’70s, I’m pretty sure we wouldn’t have given the presidency to some dude, especially a born-again Christian, who said he’d recently seen a huge, color-shifting, luminescent UFO hovering near him.” Carter's sighting (which actually occurred 7 years before he became president) was witnessed by other people. There's nothing irrational about reporting what you and other witnesses have seen, nor is it irrational to report what you alone have seen. Andersen's type of “shaming the witnesses” talk is deplorable, the type of talk engaged by those who wish to be shielded from some aspect of reality they find disturbing.

Andersen scorns those who have suspicions about evolution dogmas, but also scorns those who believe in UFO's or ancient astronauts. What sense does that make? If Darwinian assumptions are correct, we might expect that there should be life on many of the billions of planets in our galaxy. In that case we might well expect to have had extraterrestrial visitors who visited in the past or are visiting now. We may note the very arbitrary selectiveness of what Andersen deems to be irrational. Why is the ancient astronauts thesis advanced by von Daniken irrational, when the same idea was suggested in print earlier by distinguished astronomer Carl Sagan, a hero of people who call themselves rationalists?

After engaging in many an unfair characterization, Andersen then goes on to try to link these supposed examples of irrationality with the baffling election of Donald Trump, using a kind of logic insinuating that the election of 2016 was the result of seeds planted in the 1960's and 1970's. His reasoning is not convincing. When judging claims that event X helped to cause event Y, we should always look at the time between the two events. The longer the time between the events, the less likely the two are to be causally related. In this case there is some 50 years between the hippies of the 1960's and the election of 2016. Any attempt to suggest the first “planted the seeds” of the second is not believable. The “money isn't very important” hippies of the 1960's criticized the “creature comfort goals” of the moneyed establishment (to quote a phrase from a 1960's song), and would not likely have approved of Trump's election.  

The fact that Trump did not actually win the popular vote, and lost by nearly 3 million votes, is a fact inconsistent with Andersen's thesis suggesting Trump's election is a sign that America has lost its mind.  You could just as easily argue the popular vote totals were a sign of America's good judgment. 

A close examination of the dogmas held by "holier-than-thou" rationalists will show that these supposed rationalists are often people throwing rocks from inside glass houses, because such thinkers are often guilty of believing things every bit as irrational – or even more irrational – than many of the beliefs they call irrational. Such an examination will be found in past and future posts on this blog.

Saturday, August 5, 2017

A Dubious Claim to Have a Map of a Never-Observed

As I described in this post, there is a pattern I observe repeatedly in science reporting:
  1. A scientific paper will be released making modest claims that may not be particularly interesting.
  2. That paper will be hyped and exaggerated by a press release published by some institution related to the scientific paper, perhaps a particular university or scientific group.
  3. Further hype and exaggeration will be done by the popular press, which is always eager to sensationalize the scientific, because of the Internet profit that results from an increased number of users clicking on a click-bait link.
By the time the average person reads the story prompted by the scientific paper, they will be given some idea that may not at all be justified by the original paper.

I saw such a thing going on in a recent announcement of a new map called a “dark matter map.” When I saw the story announcing this on bbc.com, I was quite surprised. The bbc.com story announced, “Researchers have released the most accurate map ever produced of the dark matter in our Universe.” But how can someone have a map of dark matter locations when dark matter has never been observed? All attempts thus far to make direct observations of dark matter have failed. Dark matter doesn't even have a place in the Standard Model of Physics, and no evidence for it has turned up at the Large Hadron Collider. 

I began examining the sources of these claims. The BBC story took me to the the web site of something called the Dark Energy Survey. On that site was a press release issued by Fermilab, a major scientific organization. The press release was entitled, “Dark Energy Survey reveals most accurate measurement of dark matter structure in the universe.” That is pretty much the same as claiming to have a map of dark matter in the universe. The Fermilab press release claimed that scientists had “precisely measured the shapes of 26 million galaxies to directly map the patterns of dark matter over billions of light-years, using a technique called gravitational lensing.” It then gave a link back to the Dark Energy Survey page and some papers released on that page.

This sounded very fishy to me, because observations of gravitational lensing are not equivalent to observations of dark matter. Gravitational lensing is a strange effect produced on light rays bent by the gravity of high concentrations of matter. Such matter can be any type of matter: either normal matter or possibly some type of dark matter. As Scientific American puts it when describing gravitational lensing:

According to Albert Einstein’s general theory of relativity, mass warps space, so a large amount of matter in the foreground of a galaxy can bend its light in a way that makes it look slightly squashed. This is true whether the foreground mass is made of invisible dark matter or ordinary matter.
 
So if you are claiming to have a map of dark matter made by observing gravitational lensing, you are doing something rather like announcing that you have a map of UFO landing sites made by observing small burnt patches in the forest. Such patches might be produced by hot UFO's that are landing, but they also might be produced by ordinary lightning flashes.

What I also thought was fishy was that the Fermilab press release gave us a visual which it tells us is a map of dark matter. But the map legend on that visual does not refer to dark matter, but is instead labeled “Density of matter.”  The map is below:


To further investigate whether both the BBC and Fermilab are guilty of exaggeration or hype or dubious interpretations, I tracked down the scientific paper that is the source of these claims. The paper is here. It is entitled “Dark Energy Survey Year 1 Results: Curved-sky Weak Lensing Mass Map.” This paper has the same visual that appears in the Fermilab story and the BBC story, so it seems to be the source of their stories.

I read the abstract of the paper. There was no mention made of dark matter. I tried searching for “dark matter” in the text. The only two mentions of dark matter were incidental mentions not claiming to have observed or mapped dark matter. The first mention was this:

Briefly, three flat LCDM dark-matter-only N-body simulations were used, with 10503, 26003 and 40003 Mpc3h−3 boxes and 14003, 20483 and 20483 particles, respectively.

Simulations? That's “make believe” stuff, not an observation of dark matter or a mapping of dark matter.

The second and last mention of dark matter in the scientific paper was the following turgid prose:

Galaxies are assigned to dark matter particles and given rband absolute magnitudes based on the distribution p(d|Mr) measured from a high resolution simulation populated with galaxies using subhalo abundance matching (SHAM) (Conroy, Wechsler
& Kravtsov 2006; Reddick et al. 2013), where d is a large scale density proxy.

This is not a statement making any claim about dark matter. The scientific paper makes no claim at all to have mapped or observed dark matter. It only claims to have observed gravitational lensing and made a map of “mass distribution,” which could be any type of matter, either regular matter or dark matter. The paper did have several visuals like the visual in the Fermilab press release (which calls the visual a map of dark matter). But in the scientific paper none of those visuals was described as a map of dark matter.

Rather than claiming to be making a map of dark matter, the paper claims to be creating a map of the “mass distribution of the universe,” a more general term referring to any type of matter, either regular or dark matter. The paper states in its first sentence: One way to map the mass distribution of the Universe is by using the technique of weak gravitational lensing.” Then in its conclusion the paper states the following:

Weak lensing allows us to probe the total mass distribution in the Universe. One of the most intuitive ways to visualize and comprehend this information is through weak lensing convergence maps,or mass maps....In this paper, we construct weak lensing mass maps for the first year of Dark Energy Survey data (DES Y1) using two independent shear catalogs. METACALIBRATION and IM3SHAPE, in the redshift range 0.2 < z < 1.3 and in the region overlapping with the South Pole Telescope footprint.

This is exactly how a paper would speak if it were presenting a map of the mass distribution of the universe (its total matter that is either regular matter or dark matter), not specifically a map of dark matter.

What we have going on here by bbc.com and Fermilab seems to be shameless hype and exaggeration, which includes the inaccurate claim that a map has been made of dark matter. The bbc.com and Fermilab press coverage refers us to the Dark Energy Survey. But when the relevant scientific paper is tracked down from the site of the Dark Energy Survey, we find that the paper did not actually make any substantive claim at all about dark matter, referring to it only in two passing references. The paper does not claim to have presented a map of dark matter, but merely claims to have made maps of mass distribution (a term that means the total amount of any type of matter, whether dark or regular). A Scientific American article describes the research correctly, saying it is a finding about "the distribution of matter," and making no claim that it involves  any type of map of dark matter. 

Described by those who believe in it as something "invisible," dark matter has never been observed, and you cannot make a map of something that has never been observed. The map these stories referred to is a map of mass distribution, not specifically dark matter. But, of course, if you describe such a map as a map of dark matter, that will result in more web traffic, because it sounds like something new and exciting (scientists have been making mass distribution maps for decades).

A lesson we may draw from this episode is perhaps that exaggeration and misinterpretation of scientific papers is not something done only by more sensationalistic sites like  DailyGalaxy.com; it is also something that can be done by the most mainstream and respected science-reporting sites such as the BBC site and the Fermilab site.