Stephen Wolfram (and I) on free will

July 7, 2012 • 3:00 am

The European has an interview with Stephen Wolfram, (Many of you will know Wolfram as the designer of the famous computer program Mathematica.) The interview is mostly about progress in  computer technology and its social benefits, but there’s one question about free will.

The European: This equation of nature and technology raises a few very hard questions about the definition of life, about free will, and about intelligence. What are the mathematician’s answers to those questions?
Wolfram: Let’s talk about free will first. The problem with human free will is that deterministic stuff is going on underneath, like the chemical processes in our brain, but that we don’t seem to act in a deterministic way. People used to think that deterministic processes must result in deterministic behavior, and that belief has underpinned much of the debate about free will. It’s the reason why the science fiction robots of the 1950s often speak very logically and behave very stupidly. The main scientific discovery is that it must not be like that. We can have simple deterministic underpinnings that result in very complex and seemingly random behavior.

Computational irreducibility is a key feature of life: We cannot grasp life through a formula, but must really simulate and observe it to see what happens. That’s how we as humans end up freeing ourselves from the deterministic rules. I tend to think that the concept of computational irreducibility is probably the answer to the philosophical debates of the past two thousand years about the relationship between free will and determinism. Philosophy is always at a certain distance to human behavior, so a lot of questions really get answered by science.

This answer seems to me misguided, because deterministic underpinning that can result in “unpredictable” behavior doesn’t bear on the classical issue of free will. First of all, “deterministic underpinnings” mean that behavior (including “choice”) is still determined, even if “complex and seemingly random.”  After all, chaos theory, which is purely deterministic, yields systems whose outcomes are unpredictable if you don’t know the exact starting point. But chaos in our brain doesn’t give us free will.

As for this sentence,

I tend to think that the concept of computational irreducibility is probably the answer to the philosophical debates of the past two thousand years about the relationship between free will and determinism.

I simply don’t understand it.  It seems to mean that if we can’t reduce behavior to predictive equations, we don’t have free will.  But that doesn’t bear on the question of whether all our behavior is determined purely by our genes and our environments, and that we have no real choices.

Now my definition of “real choice”, i.e., free will, has always been this:  if we could rerun the tape of life to the moment when we made a decision, with every particle in the universe in the same place at that moment, and yet we could have made a choice different from the one we did, then that is free will.

But, as some readers have pointed out, if quantum indeterminacy obtains, then perhaps rerunning the tape could yield a different decision, and that contradicts my assertion that I don’t accept quantum effects, which are random, as part of free will.

I then modified my definition to include “if you rerun the tape of life and could have consciously made a different decision at that moment, then you have free will.” But that doesn’t work either, for quantum effects might manifest themselves by making your decision seem conscious—especially if, as some experiments show, a conscious decision is simply how our brain perceives a decision that was really made unconsciously.

So for me (and I realize this isn’t true for others), whether we have free will or not boils down to whether quantum indeterminacy holds at the level of the brain and human behavior. Some people say yes; others say no. This is a matter for empirical study. But whether or not those effects exist won’t affect the conclusions of philosophical compatibilists: folks like Dan Dennett who claim that pure determinism is compatible with free will. To compatibilists, free will is completely independent of the issue of physical “causation.”

Recognizing a cat

July 7, 2012 • 2:58 am

Elebenty gazillion readers sent me this story (with diverse links) over the last ten days.  Thanks to all of you, who are now too numerous to name. (See links at The New York Times, Wired Science, and the BBC, among others.)

The interesting results are encoded in scientific language in the abstract of a yet-unpublished manuscript by Le et al. (reference and link below). If you’re like me, the following won’t make a ton of sense:

We consider the problem of building high- level, class-specific feature detectors from only unlabeled data. For example, is it possible to learn a face detector using only unlabeled images? To answer this, we train a 9-layered locally connected sparse autoencoder with pooling and local contrast normalization on a large dataset of images (the model has 1 bil- lion connections, the dataset has 10 million 200×200 pixel images downloaded from the Internet). We train this network using model parallelism and asynchronous SGD on a clus- ter with 1,000 machines (16,000 cores) for three days. Contrary to what appears to be a widely-held intuition, our experimental re- sults reveal that it is possible to train a face detector without having to label images as containing a face or not. Control experiments show that this feature detector is robust not only to translation but also to scaling and out-of-plane rotation. We also find that the same network is sensitive to other high-level concepts such as cat faces and human bod- ies. Starting with these learned features, we trained our network to obtain 15.8% accu- racy in recognizing 20,000 object categories from ImageNet, a leap of 70% relative im- provement over the previous state-of-the-art.

At any rate, the tale is this: Google scientists interested in face-recognition software set up an immensely complicated artificial neural network, containing 16,000 computer processors, and fed it random images from YouTube.  And what came out? Cat recognition!  (Note: NOT squid recognition.)  Cats, of course, are everywhere on YouTube.

The NYT notes:

Presented with 10 million digital images found in YouTube videos, what did Google’s brain do? What millions of humans do with YouTube: looked for cats.

The neural network taught itself to recognize cats [JAC: this took 3 days], which is actually no frivolous activity. This week the researchers will present the results of their work at a conference in Edinburgh, Scotland. The Google scientists and programmers will note that while it is hardly news that the Internet is full of cat videos, the simulation nevertheless surprised them. It performed far better than any previous effort by roughly doubling its accuracy in recognizing objects in a challenging list of 20,000 distinct items. . .

To find them, the Google research team, led by the Stanford Universitycomputer scientist Andrew Y. Ng and the Google fellow Jeff Dean, used an array of 16,000 processors to create a neural network with more than one billion connections. They then fed it random thumbnails of images, one each extracted from 10 million YouTube videos.

The videos were selected randomly and that in itself is an interesting comment on what interests humans in the Internet age. However, the research is also striking. That is because the software-based neural network created by the researchers appeared to closely mirror theories developed by biologists that suggest individual neurons are trained inside the brain to detect significant objects. . .

“We never told it during the training, ‘This is a cat,’ ” said Dr. Dean, who originally helped Google design the software that lets it easily break programs into many tasks that can be computed simultaneously. “It basically invented the concept of a cat. We probably have other ones that are side views of cats.”

The Google brain assembled a dreamlike digital image of a cat by employing a hierarchy of memory locations to successively cull out general features after being exposed to millions of images. The scientists said, however, that it appeared they had developed a cybernetic cousin to what takes place in the brain’s visual cortex.

An image of a cat that a neural network taught itself to recognize. Photo by Jim Wilson for the New York Times

The BBC notes:

The work of the team stands at odds with many image-recognition techniques, which depend on telling a computer to look for specific features of a target object before any are presented to it.

By contrast, the Google machine knew nothing about the images it was to see. However, its 16,000 processing cores ran software that simulated the workings of a biological neural network with about one billion connections.

In a similar way nerves in brains are heavily interconnected and it is believed that “recognition” involves the triggering of a specific pathway through that thicket of connections.

Pathways for particular objects, people or other stimuli are thought to be built up as organisms learn about the world. Some neuroscientists speculate that parts of the human visual system become so specialised they recognise very specific subjects such as a person’s grandmother or their cat.

As millions of images were analysed by Google’s network of silicon nerves, some parts of it started to react to specific elements in those pictures.

After three days and 10 million images the network could spot a cat, even though it had never been told what one looked like.

Although the work at first seems useless, it isn’t.  We learn to recognize objects by repeated exposure to them.  And it’s always been a mystery to scientists how we’re able to form and remember images of people and friends whom we repeatedly see. (When I was younger, my father used to ask me the question, “Try to imagine a face that you’ve never seen before.” Try it—it’s not easy!) Some day face-recognition software will be everywhere: identifying you before letting you into secure facilities, helping police solve crimes, and so on.

And of course it will also give us a clue about how our brain works when recognizing faces.  We can easily clue in on human faces, but not so easily on the faces of individuals from different species, which may be nearly as distinct from each other as one human is from others.  And there’s an evolutionary reason for that: it was crucial for our group-living ancestors to recognize not only kin but other groupmates (who helped us before, and who was bad to us?), and to discriminate group-mates from potentially hostile members of outgroups.

_______

Lee, Q. V. et al. 2012. Building high-level features using large scale unsupervised learning. Manuscript at Cornell University Library. (free pdf download at link).

O Canada! Part deux

July 6, 2012 • 4:06 am

So I’m looking for dinner in Ottawa last night and find a pub that offers food.  Famished, I order a pint and a burger, and ask for the burger to be cooked rare. (I like my beef rare, ordering it “saignant,” or “bloody,” when I’m in France.)  The waiter looks at me resolutely and says, “Sorry, sir, we can’t cook it that way. According to Ontario standards, it has to be cooked between medium and medium-well.”

MEDIUM WELL!  What barbarism, but that’s the way I had to have my burger: overcooked and dry.

What are these “Ontario standards”?  I asked the waiter whether if I crossed the bridge to Quebec (just on the other side of the river), I could get a rare burger.  “Yes, of course,” she replied.  Of course, because the French (the ancestors of Québécois) are not barbarians about food, and know how beef should be cooked.

Given that every government document and sign in this entire country has to be printed in both French and English, you’d think that every province could adopt the same government standards for cooking burgers. Canadians, what gives?

The only good burger—or steak for that matter—is one that’s cooked rare, though medium rare will suffice in a pinch.

This is NOT how it should be done:

This afternoon, after the meetings are over, a few of us, guided by the estimable Larry Moran, are going in search of the best poutine in the area.  If you don’t know what that comestible is, look it up.  And when I post about it upon my return to Chicago, I don’t want any of you telling me how unhealthy it is. That’s true, but this will be only my second taste of the stuff in my lifetime.  Other participants will include Carl Zimmer and Rosie Redfield.

Poutine: the archetypal “heart attack on a plate” (or in a styrofoam box)

Guest post: America’s scientists reluctant to admit they’re nonbelievers

July 6, 2012 • 3:28 am

Elaine Ecklund continues to squeeze every drop of publishable data out of her Templeton-funded survey on religiosity among America’s scientists. As guest poster Sigmund notes below, she’s parlayed a single questionnaire into six academic papers as well as one book—a remarkable achievement in data-mining, and reason enough for Templeton to give her another $1,080,000 to continue proving that America’s scientists aren’t nearly as atheistic as everyone thinks. (Note: the conclusions of Ecklund’s work are always preordained, since she spins her data to conform to Templeton’s accommodationist line.)

In today’s guest post, Sigmund describes how Ecklund and her collaborators have squeezed the last drop of juice from her Templeton grant, publishing a paper on the reluctance of America’s scientists to answer questions relating to religion.

_____________

Elaine Ecklund finds scientists are reluctant to admit being non-religious

by Sigmund

It’s that time of the year again. The sun is shining, the birds are singing, and Elaine Howard Ecklund has returned to take another stab at the data from her 2007 Templeton -funded survey, Religion among Academic Scientists (RAAS). In the sociological equivalent to the miracle of the loaves and fishes, Ecklund has now managed to produce one academic book and, remarkably, six peer-reviewed papers from this single survey, a questionnaire designed to gauge the levels of religiosity amongst scientists at the top twenty universities in the US.

As you might guess, by this point there isn’t really much data left to mine, but the digging continues, with ever-decreasing amounts of ore.

In the new paper, ‘Missing Data in Sociological Research: An Overview of Recent Trends and an Illustration for Controversial Questions, Active Nonrespondents and Targeted Samples’, published in the journal  The American Sociologist, Ecklund and her co-author, Jeremy R. Porter from the City University of New York, approach the questionnaire from another angle. They examine the reasons why scientists may have been reluctant to answer particular questions.

In any detailed sociological survey, there will be questions left unanswered by a proportion of the participants. Sometimes the unanswered questions will be random. Other times, however, specific questions will be left unanswered at a higher frequency than chance. Termed ‘missing not at random’ (MNAR), such questions may indicate subjects of a controversial nature for the participants. Thus the non-random absence of an answer may provide a means for the researcher to gain knowledge about the views of the study participants.

The RAAS study contained questions on several subjects shown in previous studies to produce a MNAR response from some groups. These included questions about race, income, marital status and aspects of religiosity. Porter and Ecklund’s analysis reveal, however, that only questions about religiosity had answers missing at a rate higher than chance.

The questions, shown in Table 3 of the paper, included “Which of the following comes closest to your views about truth in religion?”, and “Which of the following statement comes closest to expressing what you believe about God?”, both of which had several options that could be chosen by the participant and which were unanswered by just over 10% of survey participants. Here are the data from the paper:

The question that resulted in the greatest number of missing answers was the final one shown in Table 3: “Compared to Most Americans, where would you place your RELIGIOUS views on a seven point scale?” – with answers running from ‘Extremely Liberal’ which is scored 1, to ‘Extremely Conservative’, scoring 7. 

Over 34% of scientists failed to answer this question, a figure that Porter and Ecklund found surprisingly high (although it should be noted that the initial paper revealed that over 62% of scientists were either atheists or agnostics and less than 10% said they had no doubts about God’s existence—figures that may explain the reluctance of many to answer this question.)

A well designed sociological survey should contain some degree of redundancy between questions so that the failure to answer one question can be counteracted by taking other answers into account. It was therefore possible with the RAAS survey to determine the religiosity of participants using other answered questions and compare that to the MNAR response. This technique allowed the authors to determine whether those avoiding answering the questions were religious or non-religious.

Considering that universities are amongst the least religious environments in the US, and the scientific profession is often accused of being hostile to the religious, the results were surprising.

According to Porter and Ecklund:

In terms of religiosity measures, scientists who were the least religious were the most likely to have missing data for questions on belief about God and on the comparison of their religious views to those of other Americans.

In fact it was the non-religious scientists who behaved as though their lack of religiosity was something they preferred, or needed, to keep secret. Religious scientists, on the other hand, were found to be far more likely to answer all questions related to religious beliefs and views of the bible, confounding the notion (promulgated by Ecklund her previous writings) that religious scientists are particularly worried about negative consequences from revealing their faith.

The authors speculated that the anti-atheist bigotry found in US society at large may make scientists reluctant to identify as non-religious, even within a supposedly confidential survey, due to a fear of possible negative repercussions.

Researchers find that atheists and the nonreligious are somewhat marginalized in the general population (Edgell et al. 2006). This means that scientists who are not religious may be especially unlikely to answer questions about religion for fear that the larger public might use such results against them.

 One final point worth noting is the description of exactly how Ecklund gathered the 2194 participants for the initial RAAS survey.

“The scientists included in the study were randomly selected from seven natural and social science disciplines at universities that appear on the University of Florida’s annual report of the “Top American Research Universities.“

They continue:

“Initially, the study’s PI wrote a personalized letter to each potential participant in the study that contained a $15.00 cash pre-incentive, to keep regardless of decision to participate in the survey.”

In other words, the entire study began with Ecklund stuffing 33 thousand dollars of Templeton cash into envelopes addressed to the top scientists in the US, hoping they would cooperate.

Then again, isn’t that the Templeton modus operandi—the only difference being the size of the bribe?

Nice photos from MSN

July 5, 2012 • 1:00 pm

Alert reader Steve called my attention to a “must-see photos” section of the MSN website. Well, not all of them are must-see, but a few are nice and related to biology or astronomy (click to enlarge):

First, is this “alstruism”? The stinging worker bee (a female, as all workers are) is going to die, for she’s just lost her internal organs in defense of the nest.  But it’s not genuine altruism, for by sacrificing her life, she’s perpetuating her genes, shared by her mother who will produce hundreds more sisters that carrying this same gene for defense.  Thus the loss of life accrues a genetic gain. This evolved by kin selection, not group selection!

I suspect the photograph was wrong as I recall having seen similar pictures elsewhere. Still, it’s a nice photo, and demonstrates very well how natural selection can cause one to commit suicide for the greater good of one’s genes.

SEE PHOTO HERE

July, 3, 2012: Photographer Colleen Pinski recently released this amazing image of the solar eclipse on May 20 in Albuquerque. She traveled 370 miles in a day to find the right spot and snapped this picture from 1.5 miles away. The guy walked into her frame at the perfect moment. (© Colleen Pinski/Caters News)
July 2, 2012: Just in time for the film premiere on July 3. The Mwanza Flat Headed Agama lizard bears a striking resemblance to Spider-Man and even captures his crouching pose perfectly, albeit in Kenya rather than the Big Apple. (Let’s not tell him Lizard is the villain in the latest flick.) (© Cassio Lopes/Caters News)
June 28, 2012: The ‘pig whisperer’ hypnotizes a piglet in Duelmen, Germany. Dutch veterinarian Kees Scheepens, who earned the nickname for his uncanny ability to communicate with swine, advocates for smaller-scale pig farms. ‘Let pigs be pigs,’ he says. (© Friso Gentsch/dpa/Corbis)
July 4, 2012: Disco anyone? Praying mantis in Indonesia perches in front of spiderweb covered in dew illuminated by the moonlight. (© /Caters News)
June 27, 2012: Alien-like, night-shining clouds hover over the Tibetan Plateau, as seen from the International Space Station. Meteor dust & rocket exhaust are thought to contribute to these unusual ‘noctilucent clouds.’ (Courtesy of nasaimage.org)

U.K.’s National Trust promotes creationism!

July 5, 2012 • 10:49 am

Reader chrisquartly called my attention to this post by Britain’s National Secular Society on the famous Giant’s Causeway, a formation of hexagonal basalt columns in Northern Ireland that is the remnant of an ancient volcanic eruption (see here for more information). It’s a UNESCO National Heritage site and is owned by the National Trust.  Here’s a photo (I’ve never visited but would love to):

The report says that the Causeway Visitor’s Center now includes an exhibit suggesting that the causeway formed as a result of the Biblical Flood:

Creationists believe the stones, which emerged from the sea-bed following intense volcanic and geological activity 60 million years ago, were in fact formed around 4,500 years ago as a result of Noah’s Flood.

Actually, this formation occurred 50-60 million years ago when extruded lava cooled quickly, forming the cracks that separate the tidy columns. Why is the National Trust lying about this to visitors? Creationist pressure, of course:

The National Trust had been under pressure from evangelical Christians to give equal prominence to its religious viewpoint in the new £18.5m (partly publicly funded) visitor centre at the UNESCO World Heritage Site on Northern Ireland’s north Antrim coast. . .

Wallace Thompson, chairman of the creationist Caleb Foundation said he was pleased with the inclusion of the creationist view:

“We have worked closely with the National Trust over many months with a view to ensuring that the new Causeway Visitor Centre includes an acknowledgement both of the legitimacy of the creationist position on the origins of the unique Causeway stones and of the ongoing debate around this.

“This is, as far as we are aware, a first for the National Trust anywhere in the UK, and it sets a precedent for others to follow.”

In a statement, The National Trust said that the exhibit gives recognition to the fact that, for creationists, the debate about the age of the Earth is still ongoing.

Here is the cowardly language of the National Trust exhibit, uncovered by reader Paul Braterman:

This debate has ebbed and flowed since the discovery of the Causeway to science and, historically, the Causeway became part of a global debate about how the earth’s rocks were formed.

This debate continues today for some people, who have an understanding of the formation of the earth which is different from that of current mainstream science.

Young Earth Creationists believe that the earth was created some 6000 years ago.  This is based on a specific interpretation of the Bible and in particular the account of creation in the book of Genesis.

Some people around the world, and specifically here in Northern Ireland, share this perspective.

Young Earth Creationists continue to debate questions about the age of the earth. As we have seen from the past, and understand today, perhaps the Giant’s Causeway will continue to prompt awe and wonder, and arouse debate and challenging questions for as long as visitors come to see it.

Despite the fact that they do mention “mainstream science,” this is absolutely ridiculous, and does severe discredit to the National Trust. Yeah, maybe among creationists the debate about the age of the Earth is “still ongoing,” but for the rest of the world—including all scientists and people not blinded by faith—that debate was settled decades ago.  To give it any credibility in such an exhibit is shameful.

Britain is supposed to be far less infested by creationists than is the U.S., and while fighting creationism at home, I want to keep the UK clean.  I urge readers to contact the National Trust (information here) and register a protest. I’m sure that if they get a few thousand complaints (include your title if you’re a scientist or geologist), they’ll remove the exhibit.  Well, at least I hope so.

To reach them by email, just write to enquiries@nationaltrust.org.uk  I’ve already sent them an email. Please tell them not to teach lies to visitors. I have sent the email below, and if you’re pressed for time you’re welcome to copy and paste it. It will take only 30 seconds.

It has come to my attention that you have an exhibit in the Giant’s Causeway Visitor Centre promoting the creationist view that the lava formation might have resulted from the Great Flood described in the Bible.

This is, of course, completely wrong. As all scientists know, that formation is 50-60 million years old. To teach that it’s 10,000 or so years old, and resulted from the great flood, instead of from an ancient eruption of lava, is simply teaching lies to your visitors.  Don’t you want them to know the truth about the Causeway’s formation?

Regardless of any debate among creationists about the age of the earth, there is no debate among real scientists about that age, which is about 4.6 billion years. I urge you to remove all the creationist lies about the age of the Causeway from your visitor’s centre. Please—do this in the interests of science and truth.

Thank you,

LIVE: PhD defense on climate change, mega-fauna and plants

July 5, 2012 • 4:30 am

by Matthew Cobb

While Jerry is in Canada, Greg Mayer and I are home alone at WEIT, so we get to look in Jerry’s cupboards and stuff, as well as posting. PhD student Jacquelyn Gill from the Univeristy of Wisconsin-Madison is an ace science communicator – she has a blog (NOT a website)  The Contemplative Mammoth, somewhere over there in the WordPress jungle, and she is on Twitter (@JacquelynGill),which is where I bumped into her.

*Today* 5 July, Jacquelyn takes her commitment to science communication to the highest and boldest level, by streaming her thesis defence over the interwebs! In the US (as in many countries apart from the UK), PhD thesis exams are preceded by a public lecture, explaining the science. Jacquelyn’s talk will be on-line, *live* at 1pm CDT (which I think is 6pm UK time – the rest of you will have to work it out). [JAC: it’s actually 7 pm UK time, I think]. Go here to see the presentation.

It’s on a fascinating topic that all WEIT readers will be interested in: “The Biogeography of Biotic Upheaval: Novel ecosystems and the end-Pleistocene megafaunal extinctions”.

In other words, what happened as all those giant mammals and birds disappeared at about the time humans popped up in their environments?  This is a big question, with implication not only for our understanding of the past, but also for trying to see into the future, and to understand what will happen in the coming decades through climate change and direct human activity. Jacquelyn’s research looks particularly at the role of plants in disturbed ecosystems, as indicated on this neat poster for her talk, made by Jeremy Parker.

Good luck to you Jacquelyn! WEIT readers all over the world will be watching! Other students might want to take note of this great intitiative.

Finally if you want to know the next installment, Jacquelyn will be moving to Providence, RI in August to begin a position as a Voss Postdoctoral Fellow at the Environmental Change Initiative at Brown University.