Search This Blog

Thursday 19 January 2017

Yet more on Darwinism's convenient convergences.

Sugar Gliders, Flying Squirrels, and How Evolutionists Explain Away Uncooperative Data
Cornelius Hunter

The scientific evidence contradicts evolutionary theory. Consider, for example, the problem of tracing out the mammalian evolutionary tree.

According to evolution similar species should be neighbors on the evolutionary tree. For example, the flying squirrel and sugar glider certainly are similar -- they both sport distinctive "wings" stretching from arm to leg. Shouldn't they be neighboring species? The problem is that, while they have incredible similarities, they also have big differences. Most notably, the flying squirrel is a placental and the sugar glider is a marsupial. So they must be placed far apart in the mammalian evolutionary tree. The problem in this example is that different characters, across the two species, are not congruent. Here is how evolutionists rationalize the contradiction:

Flying squirrels and sugar gliders are only distantly related. So why do they look so similar then? Their gliding "wings" and big eyes are analogous structures. Natural selection independently adapted both lineages for similar lifestyles: leaping from treetops (hence, the gliding "wings") and foraging at night (hence, the big eyes). [Emphasis added.]
This is a good example of how contradictory evidence drives evolutionists to embrace irrational just-so stories. Natural selection cannot "adapt" anything. Natural selection kills off the bad designs. It cannot influence the random mutations that must, somehow, come up with such amazing designs. This is the hard reality, but in order to rationalize the evidence, evolutionists must resort to this sort of teleological language, personifying and endowing natural selection with impossible powers. As often happens, a distinctive grammatical form -- "for similar lifestyles" -- is a dead giveaway. Natural selection becomes a designer.

This example is by no means exceptional. In fact, this sort of incongruence is rampant in biology. Evolutionists have attempted to deny it in the past, but it is undeniable. It is the rule rather than the exception. As one recent paper, entitled "Mammal madness: is the mammal tree of life not yet resolved?" admitted:

Despite the keen interest in mammals, the evolutionary history of this clade has been and remains at the center of heated scientific debates. In part, these controversies stem from the widespread occurrence of convergent morphological characters in mammals.
In addition to the morphological characters, evolutionists make extensive use of molecular sequence data using the so-called molecular clock method. This method, however, has a long history of problems. You can see here and here how the molecular clock method has failed, but an entirely different problem is the non-scientific misuse of this approach. Consider how evolutionists have misused it in the mammalian evolutionary tree problem:

Two articles in this issue address one such node, the root of the tree of living placental mammals, and come to different conclusions. The timing of the splitting event -- approximately 100 Ma based on molecular clocks -- is not in debate, at least among molecular evolutionists. Rather the question is the branching order of the three major lineages: afrotherians (e.g., elephants, manatees, hyraxes, elephant shrews, aardvarks, and tenrecs), xenarthrans (sloths, anteaters, and armadillos), and boreoeutherians (all other placentals; fig. 1).
Such overly optimistic interpretation of the molecular clock results unfortunately has a long history. Dan Graur and William Martin have showed how such overconfidence became common in evolutionary studies. They write:

We will relate a dating saga of ballooning inapplicability and snowballing error through which molecular equivalents of the 23rd October 4004 BC date have been mass-produced in the most prestigious biology journals.
Graur and Martin chronicle how a massive uncertainty was converted to, err, zero, via a sequence of machinations, including the arbitrary filtering out of data simply because they do not fit the theory:

A solution to the single-calibration conundrum would be to use multiple primary calibrations because such practices yield better results than those obtained by relying on a single point. Indeed, it was stated that "the use of multiple calibration points from the fossil record would be desirable if they were all close to the actual time of divergence." However, because no calibrations other than the 310 +/- 0 MYA value were ever used in this saga, the authors must have concluded that none exists. This is not true. Moreover, deciding whether a certain fossil is "close to the actual time of divergence" presupposes a prior knowledge of the time of divergence, which in turn will make the fossil superfluous for dating purposes.
Not only are uncooperative data discarded, but tests are altogether dropped if they don't produce the right answer:

The results indicated that 25% of the homologous protein sets in birds and mammals failed the first part of the consistency test, that is, in one out of four cases the data yielded divergence times between rodents and primates that were older than those obtained for the divergence between synapsids and diapsids. One protein yielded the absurd estimate of 2333 MYA for the human-chicken divergence event, and as an extreme outlier was discarded. For the remaining proteins, the mean bird-mammalian divergence estimate was 393 MYA with a 95% confidence interval of 471-315 MYA. In other words, the 310 MYA landmark was not recovered. Because neither condition of the consistency test was met, it was concluded that the use of the secondary calibration is unjustified.
In one example, a monumental dating uncertainty, roughly equal to the age of the universe, is magically reduced by a factor of 40:

Were calibration and derivation uncertainties taken into proper consideration, the 95% confidence interval would have turned out to be at least 40 times larger (~14.2 billion years).

Now of course there is little question that evolutionists will resolve their evolutionary tree problems. A combination of filtering the data, selecting the right method, and, of course, deciding there is nothing at all improbable about natural selection "adapting" designs in all manner of ways, can solve any problem. But at what cost? As the paper concludes, "Unfortunately, no matter how great our thirst for glimpses of the past might be, mirages contain no water."

On materialism's latest god.

How Physicists Learned to Love the Multiverse

Cornelius Hunter



Theoretical physicist Tasneem Zehra Husain has an excellent article on the multiverse in this month's Nautilus. In this age of the expert whom we must trust to give us the truth, Husain's transparent and clear explanation of some of the underlying philosophical concerns regarding the multiverse is refreshing. I only wish that her writing was more aware of the historical plenitude traditions. Many of the philosophical concerns regarding the multiverse interact heavily with, or even are mandated by, plenitude thinking. Husain makes this quite clear, and locating this thinking in the historical matrix of plenitude traditions would further enrich and elucidate her explanation of the multiverse hypothesis.

Plenitude thinking holds that everything that can exist will exist. As Aruther Lovejoy observed, it had an obvious influence on a range of thinkers since antiquity, including Bruno's infinity of worlds (read extra-terrestrials) and Leibniz's view that the species are "closely united," and "men are linked with the animals."

Though I don't suspect plenitude thinking had a direct influence on the initial development of the multiverse hypothesis, it doesn't take a physicist to see a fairly obvious connection. If everything that can exist will exist, then why should there be only one universe?

But a more interesting interaction comes in how physicists evaluate and justify the multiverse hypothesis which, after all, isn't very satisfying. With the multiverse, difficult scientific questions are answered not with clever, enlightening, solutions but with a sledgehammer. Things are the way they are because things are every possible way they could be. We are merely living in one particular universe, with one set of circumstances, so that is what we observe. But every possible set of circumstances exists out there in the multiverse. There is no profound explanation for our incredible world. No matter how complicated, no matter how unlikely, no matter how uncanny, our world is just another ho-hum universe. All outcomes exist, and all are equally likely. Nothing special here, move along.

As Princeton cosmologist Paul Steinhardt puts it, the multiverse is the "Theory of Anything," because it allows everything but explains nothing. Given this rather unsatisfying aspect of the multiverse, how can it be defended?

Enter plenitude thinking. An important theme in plenitude thinking is that there should be no arbitrary designs in nature. If everything that can exist will exist, then no particular designs will exist where others are also possible.

This has become a powerful element in evolutionary philosophies of science. As Leibniz explained, the entire, continuous, range of designs should be manifest in nature, rather than a particular, arbitrary design. That would be capricious.

This rule holds unless there is sufficient reason for it not to (Leibniz's PSR). If only one design can arise in the first place, due to some reason or technicality, then all is good -- the design is no longer viewed as arbitrary. The problem is, we can find no such reason or technicality for our universe. It seems any old universe could just as easily arise.

Plenitude thinking mandates that the designs we find in nature should fill the space of feasible designs. We should not find particular designs where others are possible. But this seems to be precisely what we find in our universe. It is a particular design where others are possible. Theoreticians have been unable to find any reason for why this design should have occurred.

If we say the universe was designed, then it is a design that is arbitrary, and that violates the Principle of Plenitude. The solution to this conundrum is the multiverse.

This is how physicists can learn to love the multiverse. Yes, it is a sledgehammer approach, but it satisfies plenitude thinking. Our universe is no longer arbitrary. Instead, the full range of universes exists out here. Husain beautifully explains this, and here is the money passage:

For decades, scientists have looked for a physical reason why the [universe's] fundamental constants should take on the values they do, but none has thus far been found. ... But to invoke design isn't very popular either, because it entails an agency that supersedes natural law. That agency must exercise choice and judgment, which -- in the absence of a rigid, perfectly balanced, and tightly constrained structure, like that of general relativity -- is necessarily arbitrary. There is something distinctly unsatisfying about the idea of there being several logically possible universes, of which only one is realized. If that were the case, as cosmologist Dennis Sciama said, you would have to think "there's [someone] who looks at this list and says 'well we're not going to have that one, and we won't have that one. We'll have that one, only that one.' "
Personally speaking, that scenario, with all its connotations of what could have been, makes me sad. Floating in my mind is a faint collage of images: forlorn children in an orphanage in some forgotten movie when one from the group is adopted; the faces of people who feverishly chased a dream, but didn't make it; thoughts of first-trimester miscarriages. All these things that almost came to life, but didn't, rankle. Unless there's a theoretical constraint ruling out all possibilities but one, the choice seems harsh and unfair.

Clearly such an arbitrary design of the universe is unacceptable. (By the way, Husain also adds the problem of evil as an associated problem: If the universe was designed, then "how are we to explain needless suffering?")

The multiverse solves all this. True, the multiverse is an unsatisfactory, sledgehammer approach. But it saves plenitude, and that is the more important consideration.

Husain's article is a thoughtful, measured explanation of how physicists today are reckoning with the multiverse hypothesis. But make no mistake, religion does the heavy lifting. The centuries-old plenitude thinking is a major theme, running all through the discourse. That, along with a sprinkling of the problem of evil, make for decisive arguments.

The multiverse is another good example of how religion drives science in ways that are far more complex than is typically understood.

post human Nanny?

Turn Over Child-Raising to a Computer?
Brendan Dixon

When our kids were small, my wife and I used a baby monitor. It was quite basic: We could hear our kids (when we left the volume up) and watch a jerky jiggle of red lights as noises ebbed and flowed. But the monitor kept us engaged with our kids. We knew better than to answer their every squeak and cry, but were concerned (and what young, new parents are not?) to know if some real danger or problem arose. The monitor covered the distance between our actual and imagined fears.

Our baby monitor made us better, not poorer, parents. If the kids needed a nap, we could, without risk, lay them down and go on to other tasks. It freed us to garden, wash the cars, do the laundry, or whatever, and remain connected to our kids. In a sense, it extended our abilities. We remained the parents. We were still "in charge." The monitor was a tool to help us do better that which we already had to do, that which was ours to do.

That describes the proper use of technology: Good technology extends us so we can be better humans. Tools that ease work, enhance connections, augment vision, and so forth enlarge us. They enable us to do better that which we need to do and that which we ought to do. Technology, however, designed to squeeze humans out by replacing our unique skills makes us less than we are. Granted, we can use some tools either way, to extend or to reduce us. But raising children is uniquely human and best not replaced by machines.

So, while we owned and benefited from a baby monitor, Mattel's new "smart baby monitor," their digital nanny dubbed Aristotle, leaves me flummoxed. Mattel, and their partner Microsoft, introduced the "Echo for Kids" at this year's Consumer Electronics Show. (Let me be clear, before going further, that, while I work for Microsoft, these are my opinions. I, in no way, represent Microsoft nor am I expressing corporate opinions on such matters.) To calm parents, Mattel and Microsoft assure that they take security and privacy seriously, following and exceeding the relevant government standards (such as, COPPA) with additional protections (since no parent wants their digital nanny to answer an innocent request with pornography).

Sidestepping Aristotle's "perky, 25-year-old kindergarten teacher" female voice, Microsoft and Mattel engineered the nanny to "read kids bedtime stories, help teens with their homework, and auto-soothe babies when they wake up wailing." Three different Artificial Intelligence engines power Aristotle. They can identify each child by voice, instruct them, and change behavior as the child grows from toddler to early-teen. Parents can also configure Aristotle; for example, to withhold that bedtime story if child fails to say "please." As one reviewer wryly comments, "never will you have to touch your child again."

Neil Postman noted our inability to measure all things or reduce them to numbers (what would it mean, for example, to say I am 31.6 percent less handsome than Bill Gates?) without loss. Neither can we reduce all things to automation without twisting the tasks into something else. Automation, even trendy, AI-driven automation, entails algorithms. Algorithms, and this includes those guiding "unsupervised" Machine Learning, encode decisions and perspectives. Someone, somewhere, somehow told the machine that this thing matters and that thing does not. Information does not arise spontaneously from matter. Digitizing divides, into zeros and ones, the flow and flux of the world. Something will get missed. Something will get cut out. Something will be valued over another thing. Neutral software does not exist.

Raising children is not a task we can or should automate. Raising a child entails training them in to live fully into that which they are: A person with gifts and abilities. Parents are responsible to inculcate values in their children. Parents, through example and training, teach children how to move through and contribute to society. And, what's more, as every parent who regrets after-the-fact uttering an inappropriate phrase, children learn at least as much by observation as by instruction. My wife and I, more than once, wondered if instruction was a waste of breath and that our children only learned by watching.

What would a child learn from a digital, even if AI-empowered, nanny? Certainly not how to be fully human. Certainly not how to behave within society. Certainly not those values and traditions and choices that make a family unique. Handing off our children to these tools reduces, not expands, our humanity. We, and our children, end up as less than we should be.


I cannot predict the future. But, I suspect, Aristotle will eventually go the way of Sony's AI dog, Aibo. Machines can be stunning, helpful tools. But, even "Deep Learning" Artificially Intelligent machines with "convolutional neural networks" are pitiful replacements for human beings. Good technology amplifies the best of us, inhibits our faults, and promotes the flourishing of the planet. Technology that replaces humans, devalues our unique gifts, and spoils where we live is not technology we should pursue.