I was listening to an NPR report today about the adaptations of the Swine Flu virus. Viruses mutate, i.e. lines of the virus “code” suddenly change, and the virus now becomes resistant to antibiotics, or is able to infect new hosts and spread faster.
That’s what people have been afraid of with the recent outbreaks: That a new “supervirus” mutation will emerge and we’ll all be dead.
On the radio program, a researcher was explaining that 10-20 years ago, an older virus that most definitely WAS related to the swine flu virus family, borrowed genes from an *external* strain of human flu virus, and the combination of those two created a newer, deadlier pathogen.
Thus we witness, in real time, the evolutionary progress of an adaptive viral machine.
I thought, there’s nobody who can grasp what’s really happening here faster than a bunch of folks like you and me who re-write and test Google ads to evolve our businesses.
So let’s dig a level deeper to see what just happened: An older virus that WAS related to swine flu copied lines of code from a different virus.
Google advertisers do that all the time. We do it like this:
You’re advertising on the keyword “business systems.” You have an ad that says:
Small Business Systems
Learn the art & science of creating
effective systems/processes. Report
www.BoxTheoryGold.com
You’re looking for some inspiration so you check out the ads for Brazilian Jiu Jitsu and you like this one:
Learn Brazilian Jiu Jitsu
Master the Art of Grappling With
The Most Complete Video Ever Made
www.BjjVideoEncyclopedia.com
And you create a new ad that says:
The Art of Small Business
Master the art & science of Systems
The Most Complete Video Ever Made
www.BoxTheoryGold.com/mastery
On NPR the guy said, “There’s no way to predict which new strain will suddenly go viral and kill 1000 people.”
BINGO. Don’t we all know that from our own personal experience? We come up with all these crazy ideas and we never know which one is going to hit pay dirt.
Sometimes what the people like to click on surprises the heck out of ya. You never know until you test. The evolution of your advertising leads you in fascinating, unexpected directions.
Notice that in the evolution of this ad we also had to go from offering a report to “the most complete video ever made.” Yeah, making the video is a tall order, but if it brings you a million dollar client, who’s gonna complain?
Now I want to point out a much less obvious point, something that was even hinted at in the NPR report but never explicitly stated:
The mutation of viruses is clearly a calculated process that is somehow intelligent and definitely purposeful.
It is NOT random.
Let me explain exactly what I mean by that.
There’s a “textbook version” of evolution that you get in biology books, it says: “Evolution is a process of random mutation and natural selection.” Almost every biology book, almost every website you find, says that accidental copying errors in DNA occasionally result in an improvement and eventually the virus or the animal evolves in a positive direction.
Let’s test the “accidental copying errors” theory with Google ads. We start out with this ad:
Simple Self Defense
For Ordinary People
Easy Personal Protection Training
www.tftgroup.com
Response: 0.8%
Simple Self Defense
For Ordinary People
Fast Personal Protection Training
www.tftgroup.com
Response: 1.3%
The difference is that we mutated one word: We changed “Easy” to “Fast” and went from 0.8% to 1.3%. We deliberately changed one adjective to a different one, which proved better.
Now if we try RANDOM mutation – the “accidental copying error” that the biology textbooks talk about – after one mutation it looks like this:
Simple Self Defense
For Ordinary People
Easy Personal Protect0on Training
www.tftgroup.com
The mutant letter is in the word “˜protection,’ and it has a 0 instead of an i.
After five mutations it looks like this:
Simple Self Def4nse
For Ordinary Peopla
Easy Personal Protect0ov Traininf
www.tftgroup.com
After ten mutations it looks like this:
Simple SPlf Def4nse
For Ordinary Geopla
Emsy Personal Protect0ov Traininf
ww8.tftgroup.com
Now you can always start over with the original ad and try again, but go ahead and see for yourself (I have a “random mutation generator” at www.randommutation.com) – if anybody can use random mutations to improve their Google ads I’ll happily pay you a $1000 bounty.
If you use only random mutations to evolve your ads…. you’ll bankrupt yourself before you get a single winner. I’ve never found anybody who’s willing to take me up on my $1000 wager.
So what’s my point?
My point is: Random Mutation is not how evolution works. Your own personal experience tells you that.
Everybody who is involved in the daily evolution of anything sees that evolution is not a completely random event that somehow gets lucky, it’s an engineered process. Success often takes a brilliant borrowing of an idea from someone else, or the best guess you can muster. Mindless accident will only kill you faster.
–> This may be one reason why some of your friends assume that you’re successful in business because you “got lucky.” They truly believe that anybody’s success is just a string of lucky accidents; after all, it’s what they were taught in school, isn’t it?
(I heard the famous atheist Richard Dawkins on the radio a few years ago, and he said the origin of life was just “a happy chemical accident.” Hey wait a minute… if Isaac Newton had accepted that kind of non-explanation as an answer, how would he have ever discovered gravity?)
If you are seriously interested in the biological side of this, there is an absolutely fascinating paper by James Shapiro from the University of Chicago called “A 21st Century View of Evolution” which documents in detail how organisms re-program their own DNA to adapt to a changing environment.
A protozoa in distress splices its own DNA into 100,000 pieces and re-arranges them to make a new, more competitive protozoa. (Think about that for a minute. WOW. Nothing in the computer programming world even begins to approach that level of sophistication.)
Tim Ferris hit a winning mutation jackpot when he came up with the book title “The Four Hour Work Week” which he originally tested in a Google ad. The book is now a New York Times bestseller.
His success – and your success – is not a ‘happy chemical accident.’ It’s engineered to happen, it’s designed to happen, and it’s intentional. Note that he had to completely re-organize his book around this theme too. Evolution doesn’t happen for free.
And like I said, nobody experiences how evolution *really* works than people like you and me who evolve products, processes, marketing systems and businesses every single day. And nobody understands the thrill of hitting pay dirt with a new mutation that goes viral and slaughters the competition, than you and me.
Even the constantly mutating, self-adapting swine flu is proof that inside every success story is an intelligent mechanism that’s constantly, tirelessly searching for a superior combination of code. Even the lowly swine flu virus seeks world domination.
You want your business to go viral because you’re obeying the core directive that’s in the DNA of all living things. You have selfish genes and that’s good…. and those genes are intentional and intelligent. Intelligence doesn’t come from nothing. All our yearnings to grow, reproduce and succeed speak of an original, ingenious process set in place long ago by Someone who is Very Wise.
Perry Marshall
Further reading:
James Shapiro’s “A 21st Century View of Evolution”; and a more recent work: http://shapiro.bsd.uchicago.edu/Shapiro.2005.Gene.pdf
My DNA lectures at Lucent Technologies: “Information Theory, DNA and the Case for Intelligent Evolution”
Jessica Bischof’s ebook on preventing Swine Flu
Share This Post
About the Author
Perry Marshall has launched two revolutions in sales and marketing. In Pay-Per-Click advertising, he pioneered best practices and wrote the world's best selling book on Google advertising. And he's driven the 80/20 Principle deeper than any other author, creating a new movement in business.
He is referenced across the Internet and by Harvard Business Review, The New York Times, INC and Forbes Magazine.
45 Comments on “Swine Flu Virus Mutations and the Evolution of Google Ads”
I wonder if you are not setting up a Darwinian “straw man” to destroy. I have not read Darwin and do not know his exact words, but Darwin’s not knowing much of the genetic transmission process (e.g., his not knowing of DNA) how can you say his “random mutations” are not your higher-order changes? It seems to me that your conjecture is just a semantic cognate of Darwin’s unknown process of genetic change. Certainly the concept of generating many “trial” changes and then selecting the most felicitous adaptation seems to be the essence of both Darwin’s and your explanations.
No Darwinian straw man here at all. Vast swaths of biology literature state (without proof) that random mutations of DNA base pairs filtered by natural selection produce new species. Please refer to James A. Shapiro, “A 21st Century View of Evolution”: http://shapiro.bsd.uchicago.edu/Shapiro.2005.Gene.pdf for a revised understanding of how evolution happens. Note Shapiro’s emphatic statements that random changes to DNA are not the way evolution proceeds.
Perhaps one could cite any evidence for a beneficial mutation, as far as I know not one has been identified yet (sickle cell anaemia is hardly beneficial)
The problem is that natural selection requires the generation of ‘new’ information in vast quantities, for which there is just no evidence. pseudo beneficial mutations, like that which causes antibiotic resistance in strpotococcus aureliis n(MRSA) are actually a loss of data,
There is also the problem of the ‘cost’ of selection, in that, for the winner to be ‘selected’, the loosers must die, this will very quickly put the organism below sustainability levels and it will die out.
Natural selection is a wonderful urban myth, and will be seen as the greatest scientific hoax of all time….
I’ll reply in more detail later, but I’m sure you agree that no human being was created by a random accidental event. It was a series of events. Millions and millions of events that eventually turned up with human beings.
The first computer program ever had how many lines of code?
Patrick,
Don’t worry. I’m not discussing this because I want to change Perry’s mind. I just like these kind of discussions and Perry luckily is somebody that doesn’t shy away from a discussion like this.
As to the subject: There seems to be a need for reproducing evolution through random mutations in order to believe that it is random mutations that drive evolution.
I have a question for Perry:
We seem to agree on the point of all the mechanisms that are in place today work towards not allowing bad mutations to have much of a chance to continue. We do know that they sometimes have a chance as sometimes deficiencies do happen.
But all that engineering seems to happen after the mutations. That’s what the shakiro paper also seems to describe. What happens after the mutations happen. It doesn’t describe what triggered it.
You´re looking at a stone flying through the air and by studying the path of that stone through the air, you notice it seems to follow some kind of engineered path as it is a very consistant path,.. as if it was planned to follow that path. But ofcourse, the path is dicated by the laws of gravity, nobody is steering that stone. It’s just under the influence of the laws of gravity.
The real question of why the stone is there on that path can only be answered if you know how and why it was thrown or perhaps it just fell of a mountain, or perhaps a horse running made it fly up in the air, or…. well many things. It could be anything that caused that stone to be moving through the air. The cause could be just about anything. In other words, a random cause.
Would you agree that even though the stone is following an “engineered” path, the cause is still random?
Peter,
I can only encourage you to re-read Shapiro’s paper “A 21st century view of evolution: genome system architecture, repetitive DNA, and natural genetic engineering.” You have misrepresented what Shapiro has said.
The whole point of Shapiro’s work is that the mutations themselves are highly engineered events. That is precisely what he means by his term “Natural Genetic Engineering.” I quote:
“Major evolutionary steps occur by DNA rearrangements carried out by sophisticated cellular natural genetic engineering systems operating non-randomly.”
If you’re looking for analogies, a stone flying through the air is not one that fits. A much closer analogy would be: You discover a computer program that not only has extremely sophisticated error-correction mechanisms which are quite similar to those you find in a DVD player, but the computer program also re-arranges thousands of lines of code on the fly to adapt to viruses and new operating systems.
Did that computer program come from some random bit of noise floating around in the universe? Or was it written by an incredibly talented engineer?
I for one know of no computer programs that were created by a random accidental event.
Perry
Hey Perry,
Thanks for the answers. I’m pleased you wrote back.
You have a good point about moral obligation. You’re probably right, that’s exactly why some fear religion.
As I was growing up, I guess I never saw my religion as directly related to my moral code (though, now that I think about it, that’s exactly what it was supposed to be for). So when my religion didn’t make logical sense anymore, I guess I didn’t feel like I was losing any part of my character.
Your rhetorical question about why some people fear intelligent design is really only asking half the question.
The real question is: Why does intelligent design vs evolution upset so many people?
The answer for atheists: intelligent design implies a challenge to their moral code, or maybe their independence.
But for Christians, any evidence of evolution challenges the foundation of their faith, and in the absence of proof for their faith (which we’ve said, by definition, cannot exist) they must destroy any proof which opposes it.
The same question which you pose to atheists “Why does this bother your moral code so much?” I could pose to Christians “Why does this bother your faith so much?”. And a similar question to evangelists: “Why is it difficult to accept that others believe differently than you?
You don’t have to answer this time, unless you feel like it. You’ve got my respect, and my gratitude.
Take care,
Patrick
————————————-
Peter,
Dude. Let it go. We’ve all had our fun.
Perry’s not going to change his mind or decide he’s wrong. Neither are you. If for no other reason, then obey the #1 rule of sales: you can’t change anyone’s mind by arguing with them.
— Patrick
Patrick,
Let me slightly re-frame your question: Why does naturalism vs. design upset so many people? (I don’t see evolution and design as being opposites; my observation is that everything that we know the origin of that does evolve, is designed.)
It’s an upsetting question because it strikes at the heart of the issue of whether the universe and human existence has purpose or obligation.
If the universe is purposeless and everything we see here is the result of blind chance and laws of nature, then that carries moral implications.
If the universe is purposeful and everything we see here is the result of an intentional act of any kind of spiritual entity, then that carries moral implications.
I think that religion that does not speak to moral obligations is pretty useless.
Everyone knows that religious people who don’t respect their own moral obligations are hypocrites.
Final thought: I don’t see “faith” as being a set of beliefs which by definition can have no proof. Not at all. I see faith as hypothesis that eventually leads to proof or at least some sort of confirmation and supporting experience. If that sounds like the scientific method to you, then you’re exactly right. In other places (for example http://www.perrymarshall.com/merry-christmas-2008/ )
I make the case that modern science owes its very existence to faith and theology. Because science rests on the presupposition that the world is ruled by fixed discoverable laws and that was originally a religious idea.
I similarly see the living of my personal faith as being more similar to that than different. It produces testable hypotheses about how DNA functions (which I lightly allude to on this very web page). I can live my life by moral principles and see the difference compared to those who live by different principles. And even the effectiveness of prayer itself is something that can be tested based on outcomes.
Perry
P.S.: Regarding your reply to Peter, I could be swayed by evidence, and have been asking for the same.
Actually about 5 years ago I was initially persuaded that evolution may in fact be driven by randomness; that all you needed was Natural Selection to kill off the losers from the random mutations and you would naturally have forward progress.
But as I investigated, I found that there is no actual evidence to support this at all. I was somewhat surprised to find this to be so. It is a VERY subtle point; most people don’t really even know what they mean when they say ‘random mutation.’ That little word “random” just gets snuck in there and people assume all is well.
Imagine my surprise in discovering that although thousands or even tens of thousands of books say that evolution is driven by “random mutations” or “random variation”, I could not find even one scientific paper that backed that claim with experimental data.
Instead what I found was people like Goldschmidt and Dobzhansky, who lab-tested this theory for decades (mutating fruit fly and moth DNA with radiation) and got absolutely nowhere.
As you so well pointed out with your own links and descriptions, organisms evolve by highly calculated, very precise re-arrangements of specific genes. This is what Shapiro calls “cellular genetic engineering.” This process isn’t the least bit random, and in fact the genome devotes considerable resources to preventing random mutations.
I’ve been debating this subject online for 4 years now and I’ve never once been presented with positive evidence for random mutation as a driver of evolution. This is always assumed, never proved.
Peter said that asking for proof of this is like asking for proof that 1+1=2. I can understand why he has always taken this for granted. Random Mutation is believed by many to be self-evident and not even requiring proof.
But I am asking for proof. And I want to be clear that I was 100% sincere in my request to Peter. My mind can be changed with factual information. If presented with empirical evidence and scientific studies I will graciously consider it.
Sorry Perry, but what you´re asking is the same thing as asking to prove that 1 + 1 is 2. When somebody asks that, there’s no convincing possible.
But that’s ok. :)
There are people that think it’s proven that we didn’t go to the moon and that it’s all a fake. To them it’s proven that we never went to the moon. It’s the same thing with them,.. if you explain why their logic is wrong, they ignore it (like you´re doing) and return to their own evidence.
The only way to prove people went to the moon is to take them there and show what was left behind in the early 70’s.
http://en.wikipedia.org/wiki/Peter_and_Rosemary_Grant
“Peter and Rosemary Grant are distinguished for their remarkable long-term studies demonstrating evolution in action in Galápagos finches. They have demonstrated how very rapid changes in body and beak size in response to changes in the food supply are driven by natural selection. They have also elucidated the mechanisms by which new species arise and how genetic diversity is maintained in natural populations. The work of the Grants has had a seminal influence in the fields of population biology, evolution and ecology.”
I’ll keep my reply short Perry,…. It’s the error correction that results in the non random result. The mutations on DNA level however, are still random.
In Shapiro’s paper, he says, “Major evolutionary steps occur by DNA rearrangements carried out by sophisticated cellular natural genetic engineering systems operating non-randomly” and he backs this up with specific examples from his research.
I’ll keep my reply short, Peter…. please likewise back up your own claim with evidence.
Peter,
If you’re saying the reason some people are taller than others is because of random mutations to DNA, it would be the first time I’ve ever heard that theory. From what I understand this is determined by specific factors of Mendelian genetics.
The number of characters in Google ads follows a bell curve distribution too, but that does not mean the source of those ads is random noise.
This is not the first time that I’ve heard a hypothesis that an error correction mechanism accidentally emerged from the errors themselves. I would be interested in anyone who can offer an empirical example of something like that actually happening.
From all that I know about engineering communication theory, the only thing noise ever produces is noise. Noise does not produce words or language, and certainly does not create error correction systems.
I’ve yet to meet anyone who can demonstrate otherwise. My invitation for anyone to present some examples of that is still open.
~
Patrick:
1) On your first point, I think we are both actually saying the same thing. I gave an example of two parent ads giving birth to a new, innovative child ad, and I said, this is highly similar to how species adapt.
2) And yes, the substitutions must be made on the WORD and PHRASE level, not at the level of single characters. Which underscores my point that evolution is not random, because randomness by definition operates at the lowest level. Changing words and following rules of syntax is what organisms do, and that is a high-level function.
2 1/2) Mutating 1 letter out of 120 in contrast to millions of genes is a valuable point and I’m glad you brought this up.
The DNA for even the simplest cell is vastly more complex and sophisticated than a Google ad. So what I am saying is this: If random mutation can’t evolve a tiny Google ad – if even in a simple PPC ad, you have to make intelligent word substitutions in order to get an improvement – then how much more should this be true of an intricate genetic program?
If you read Shapiro’s paper, it backs both what you are saying and what I am saying. That genetic adaptations are intelligent substitutions, not random mutations. The proof of this is readily available in the literature.
To your last point about autonomy and choice: I think many people resist the substantial amount of evidence for design in the universe for the sole reason that it implies that the universe also comes with moral responsibilities, obligations and/or implications that they do not like.
Some people simply prefer the materialistic interpretation, despite its inability to explain the origin of life for example. The fact that information is a top-down phenomenon is, all by itself, proof that the bottom-up, materialism / reductionism interpretation falls short.
Finally, studies show that small business owners are more likely to be religious than the general population. I don’t see religion as taking away my independence and neither do many others.
I do see religion as saying that yes, there is an objective reality that if somebody steals from you or me, it really is objectively WRONG and not merely the violation of some arbitrary convention that some humans invented to wield power over other humans.
Sincerely,
Perry
Perry,
While I don’t share your faith, your maintaining it is one of the reasons I listen to you.
Even so, your Adwords analogy has 2 and 1/2 flaws.
I’ll get to them in a moment. First, let’s agree on how biology defines genetic variation:
1.) Genetic variation occurs after fertilization, when the parent chromosomes meet for the first time.
2.) The chromosomes unravel and elongate
4.) The parent chromosomes wrap around each other like tangled spaghetti
5.) Chunks of DNA (genes) swap places resulting in genetic variation.
Here’s a picture: http://evolution.berkeley.edu/evolibrary/article/4_0_0/evo_17
Single-celled and viral mutation happens in a similar way, except the DNA (or viral RNA) makes a pretzel with itself and genes swap around.
So, yes, there’s a degree of randomness, but the randomness only reshuffles the genes which previously existed.
That brings up the first error in your Adwords analogy. Natural mutations occur with genes from inside the parent genome, not new letters from outside. So adding random characters is out. We can only draw from inside the ad.
Second, we’re swapping genes, not individual codons. There are markers within DNA which indicate where genes start and end, and that’s where the switches happen.
So, really, we should be swapping words, not letters.
If we want to acknowledge that sexual reproduction is superior to asexual reproduction, then really we should be swapping WORDS from TWO PARENT ADS to create an offspring ad. Now a functional ad is more likely. A successful ad resulting from two highly successful ads is even more likely.
Offer me those terms and I’ll take a spot in your next Bobsled run over the $1000. (Duh, who wouldn’t…)
Unrelated, is the last 1/2 flaw (for a total of 2 and 1/2 ;-) that you’re mutating 1 letter out of 120. Genomes have millions of genes. So the analogy breaks down there.
But wait! Where are my studies which back all this up? I don’t know. Ask Berkeley. As with many things in our world (like calculus proofs), I’m willing to accept that someone has done that research so I don’t have to. (As your website proves, many people feel that way about Adwords.)
I would distinguish my trust from faith, because I know the proof is available if I want to see it. By definition, faith demands acceptance without ever seeing proof.
Lastly, I think people are afraid of “intelligent design” because it is the first step down a very long, slippery slope, at the end of which is a belief system which removes the autonomy and choice from our lives.
You’re dealing with people who want to be entrepreneurs. We’re willing to work until our bodies fail us, with no safety net, for as long as it takes, if it means we can be independent. As long as religion seems antithetical to that, we’ll resist it.
— Patrick
I suppose you are trying to say that this virus was engineered because of its rapid change? If so, I highly recommend this article: http://www.sott.net/articles/show/183333-Connecting-the-Dots-A-Pandemic-Distracts-as-the-World-Government-Picks-a-Fight
Perry,
I think you´re pretty much confirming that what drives the mutations is a random process. You´re adding to that that cells however, try to not allow any mutations but that they´re not 100% successful at that and that that is what creates improvements (and also defects.)
If you would do a study of the distribution of for example shapes of human beings, you’ll find that most have an average shape, and that the further from that average shape you get, the less people you will find that have a body like that. That’s a typical description of a normal chance distribution, which is logical: Mutations that look a lot like the original will be easier to pass through the cell error correction processes than a mutation that’s very different. However, over thousands and thousands of years, the difference between the original and the current version is much bigger. Every time a small step also ends up in a big step.
Now if you look at it at different levels then you’ll find that mutations after all the engineering, is still a random process. So mutations at DNA level are random, but mutations at for example a whole body level, are also random. That’s why you see the distribution of body shapes is a random distribution.
The same applies for other properties (intelligence, strength, height, etc. etc.) they all are distributed the same way.
It all proves that the whole process is random, from the cell level mutations to complete changes in whole bodies.
I agree with you that the error correction exists. It’s a necessity even because it improves survivability. But that’s the whole point of Darwinian Evolution: Properties that improve survivability have a better chance to stay. Thus error correction is something that has evolved as well. It wasn’t put in there by somebody or something. (If you believe in God and believe that God put it in there, then I’ll go as far as agreeing that God has made the laws of nature, knowing (He knows everything) that evolution based on random mutation is the best way to go.)
As to your random mutation generator, it doesn’t do what nature does, therefore it doesn’t proof anything.
Your random mutation generator tries to create a better version of an existing body that was created through evolution. You then go way back in time when the very first random mutations created the first improvements, use that level of simplicity to mutate that (advanced) body into a new body and are surprised that that doesn’t work.
I think it’s pretty obvious that that doesn’t work.
Almost,. :)
Just because there is error correction, doesn’t mean that those mutations that do pass through, are deliberately chosen by somebody or something.
Your oldest child doesn’t look exactly the same as your youngest child, even though they were created from the exact same 2 parents. That’s pure chance at work, nothing more.
If you don’t agree with this, try explaining to parents that have a child with Down syndrome that their child was intelligently engineered.
It’s all pure chance (which is the same thing as random)
The combination of two parents’ genes is random in a limited sense, but it exactly obeys the laws of genetics in the larger sense.
This is wholly different from “random mutation” which is the origin of all manner of birth defects, like Down Syndrome. So yes you are right, those mutations that pass through are NOT necessarily chosen by somebody or something. The genome guards militantly against random mutations so that things like Down syndrome do not occur, but as we know with computer systems, error correction systems are not 100% failure proof. Sometimes corrupted code slips through.
Traditional Darwinian evolution gives the credit for the magnificent forms of biology to random mutation, when in fact random mutations produce almost nothing but birth defects, extinction and death. What a sad irony. It is a crime against the enterprise of science to make this atrocious reversal.
It is not all pure chance, and furthermore, there is no empirical evidence to suggest that it is pure chance. Notice that throughout this thread there have been numerous objections but no one has produced any experiments that demonstrate that evolution is driven by random mutation. Which is reinforced by my original analogy, which is that nobody can manage to re-write Google ads successfully with a Random Mutation generator – no matter how much natural selection is available.
As I said, I’ve been watching for evidence that random mutation drives evolution for 5 years now and there is none. Always assumed, and taught ’round the world, but never proven.
Perry,
I think it’s obvious that most random mutations are NOT applied. The Shapiro paper shows just that.
That’s right, Peter. The genome militantly guards against most “possible” mutations, and only executes deliberate transpositions that it predicts will be beneficial.
Which means the traditional Darwinian theory can’t be right. Shapiro’s theory of evolution is superficially similar – because the end result is the same, and because natural selection operates the same way. But underneath it is wholly different from Darwin’s theory. It’s driven by cellular engineering, not random copying errors.
Perry just reminded me of an article I just read.
http://io9.com/5227470/a-drug-to-re%20awaken-ancient-human-genes-and-fight-hiv
Scientists are looking at ancient sections of our DNA code. They have found that a mutation in those sections may have blocked a chemical which they know blocks transmission of the HIV virus. Who knows what other turned off sections are hidden. Again there is a system. It makes sense that DNA invests heavily into error correction, look what happens when an important gene gets turned off.
I was more thinking of a simpler experiment where the only thing that happens is to see if an ad that gets randomly mutated + natural selection (i.e., anything that’s not a real word gets converted back to the previous word) will eventually end up with an Ad that has only that single word changed from “Easy” to “Fast”.
To speed that process up, we will reduce the time it takes to mutate single words, by just randomly changing words, using a database of all english words available. To simplify that a bit more we can also use a reduced word set. This just to speed things up, it doesn’t change the validity of the experiment.
But,…. somehow I think that you don’t agree with these settings.. :)
You´re making a fundamental mistake with your random mutation simulator. Assuming that all mutations are automatically applied is just not correct. You´re taking natural selection out of the equation. I don’t think we’ll ever agree on that point. But that’s ok. We can just agree to not agree on that point and leave it at that.
Peter,
Yes of course all mutations are automatically applied. All mutated ads have to be displayed, just as all offspring that are conceived must attempt to survive. Not only do they have to survive after they’re born, they have to survive the birthing process. Darwinian evolution doesn’t get to have help from a “dictionary” and you don’t either. Of course that would save time but it also sidesteps the point of actually simulating Darwinian evolution. Darwinian evolution has no way of knowing in advance what has a chance of working, and nobody gets to cheat by trying to help it along.
Perry,
This is exactly why I don’t agree with your random mutation simulator. It doesn’t follow the rules of evolution, all it does is random mutations. But that’s not how evolution works.
Random mutations happen all the time, but natural selection makes sure that the good mutations stay. Error correction is part of natural selection.
If I can’t add natural selection to the simulator, then that’s not fair.
You can easily add natural selection to the simulator. Take any output obtained from the Random Mutation generator and put it in front of people as an ad and see if it gets clicked on.
There are certainly lower-tier PPC systems that would allow you to do this unimpeded. You could easily do this with banner ads that are pulled from your own server, for example. Just randomly mutate the JPG files and serve them.
For the sake of this experiment, you’re welcome to go ahead and use Google. Let’s just make sure everyone acknowledges in advance that Google is a carefully engineered, DESIGNED system which automatically weeds out all kinds of things INTENTIONALLY. If Google won’t let mis-spelled words through then it’s not an accurate simulation of Darwinian evolution.
But I’m still not worried because even with all of Google’s help I don’t think you can re-write a better ad with a random mutation generator.
Like I said, by all means try it using Google AdWords and report back to us with the results.
Perry
Perry,
I’ll read the Shapiro paper as you suggest. If it shows the strong empirical evidence that you believe it does, I’ll change my opinion.
But I’m still prepared to fault your logic. I certainly won’t change my view over your forced and wildly inaccurate metaphor that compares genetic mutation to adwords authorship.
Regarding your points about purpose and design, I’m a little skittish about where you seem to be headed with your argument.
Do you believe it’s possible to postulate the role of purpose in evolution without assuming the existence of a higher being?
I feel a “prime mover” argument lurks behind your observations. Let me be sure I follow your logic.
Where would purpose, in the true sense of the word, reside? Purpose is intention, and intention implies motivation that can occur only in sentient beings.
Viruses are not sentient beings. So the purpose you attribute to the genetic mutation of viruses must come from outside of them.
Where would purpose come from if not from a “prime mover” that governs the laws of nature?
Or are you saying that what you call “purpose” is much less than we normally think of it. Are you saying it could somehow be hardwired into the inherited behavior of organisms — even ones as simple as viruses?
When a plant grows toward the sun, is it evidence of purpose in or outside of the plant? Does it point to the existence of a prime mover?
I think a plant’s motion toward sunlight is evidence only that the plant obeys physical, chemical and biological laws at a cellular level.
If you want to believe ideas for which there’s no evidence, go ahead. Just call it faith.
Let’s not kid ourselves that what you think you “know” by faith is as solid as knowledge that stands up to scientific method.
Dave
Dave,
Read Shapiro’s paper first, then we can discuss.
Perry,
That document of Shapiro does not mention anywhere the 2 words “Intelligent design”. It’s says that cells have natural genetic engineering capabilities, which I completely agree with. But that doesn’t mean that there is a predetermined goal (which is what intelligent design implies). It does mean that cells are capable to ignore useless random mutations and even repair bad mutations to go back to a previous state that works better. That’s something that’s even needed because without error correction all random mutations have equal chances and that’s not very beneficial to success.
Still it can be argued that the original first cell(s) (I think there were more than 1) also evolved to have these error correction capabilities because it just improves their survival chances.
A living system like a complete human being can’t stay a live if it’s cells constantly mutate. It’s required that they protect them selves from too many mutations.
Here’s the thing, as I want those 1000 dollars,. :) I’ll write a program that changes 1 ad to another through random mutations, but with error correction rules applied so that we get a more correct model of evolution that still is driven by random mutations.
Peter,
More power to ya going for the $1000. But you don’t get to cheat. Externally added error correction rules are cheating. All you get to do is create the mutated ad and throw it in the lake and see if it can swim.
Google’s staff reviews ads for content, so in a sense, that is cheating too. That would be like lizards laying eggs and some guy saying “let’s just remove those from the gene pool, I don’t think they’ll do very well” and putting them in his pocket.
If you have an advertising system that does NOT check spelling or have a review system in place, and the randomly mutated ads win or lose purely on the basis of CTR, then you have something highly analogous to random mutation + natural selection.
I encourage you to try this and see what happens.
Perry Marshall
Peter,
In researching this topic for 5 years I have not encountered any direct evidence that random mutations drive evolution. I insist that this is an urban myth and that when pressed to actually prove this is true, everyone comes up empty.
I am fully open to being corrected. A satisfactory answer to this question requires actual empirical evidence, not assertions.
Shapiro has discovered that organisms devote considerable resources to preventing accidental random mutations – via mechanisms very similar to the error correction mechanism and redundancy features in computer systems, DVD players and the Internet.
Meanwhile these same organisms do mutate in very controlled, specific, algorithmic ways, when subjected to stress. Barbara McClintock discovered this in 1944 and won the Nobel Prize for her work in 1983. She called it “Genetic Recombination” and Shapiro’s work is a continuation of her original discoveries. Fascinating stuff.
The genetic code is every bit as complex and subject to just as many rules as human language. That’s why random mutations damage DNA just as much as they damage English or Chinese. Random mutation = devolution, extinction and death.
I submit to you that the “evolution through random mutation” hypothesis is one the greatest and most subtle half-truths in all of modern science.
Again, I invite whatever data you have. And I invite you to read Shapiro’s paper.
Perry Marshall
“I agree with you, I think that some version of evolution does match what we see in the world. But the larger point is, evolution itself is always driven by a goal-seeking program.”
If you change the word “program” to “algorithm” I will agree with it.
Algorithm actually just means some kind of loop backing logic that works based on certain laws. In this case laws of nature and that’s actually just mathematics.
If you want to believe somebody wrote the laws of nature, then that’s fine by me,. :) That I don’t want to argue about. But the fact that random mutations drive evolution you can’t contest. That’s just a fact.
The mutation from “Easy” to “Fast” is also a possible random mutation. It’s not like random mutations happen 1 letter at the time. They happen in great numbers at the time and most of them are useless, meaning they have neither positive or negative impacts.
But over time, a number of random mutation can result in something new that’s actual beneficial.
There is no conscience intelligence that plans mutations.
“Q: Why is the idea that evolution is purposeful offensive to some?”
I wouldn’t consider it offensive, but this question can be mutated (pun intended :) ) to: “Why is the idea that evolution is purely the result of random mutations, offensive to some?”
The thing is,… The idea that a change from “Easy” to “Fast” is always a planned mutation and can not happen randomly, shows a lack of understanding of how random mutations work. Random mutations do not automatically imply nonsense results.
Andy,
You said, “If so, then it may be that one could improve one’s Google ads, assuming one was working on geological time scales of millions of years and had several trillions of dollars to spend on adwords.”
http://en.wikipedia.org/wiki/Infinite_monkey_theorem
From it –
“it produced this partial line from Henry IV, Part 2, reporting that it took 2,737,850 million billion billion billion monkey-years to reach 24 matching characters:
RUMOUR. Open your ears; 9r”5j5&?OWTY Z0d… ”
Although that was trying to match specific text, mutating an ad into a better performing ad would basically never happen during any realistic concept of time unless the mutation only required a few character changes.
Q: Why is the idea that evolution is purposeful offensive to some?
A (from Richard Dawkins): (A)lthough atheism might have been logically tenable before Darwin, Darwin made it possible to be an intellectually fulfilled atheist.
Purposeful evolution removes that option.
(Though I would quibble with the word “purposeful,” which implies “purpose” to be a characteristic possessed by and inherent in nature (see the recent work, “Reinventing the Sacred for the “Emergent View” which posits creativity as an inherent property of the natural universe.))
I prefer the term “teleological” which implies a purpose given to evolution.
Dave,
I submit to you that:
The idea that accidental recombination of genes is the source of evolution is, itself, an article of faith.
It is believed by many scientists, but this belief not supported by empirical evidence. I encourage you to read Shapiro’s paper and judge his findings for yourself.
Q: Why is the idea that evolution is purposeful offensive to some?
Perry,
I agree with nearly everything you said in your post except your argument regarding evolution.
It seems to me that you’re misusing an adwords metaphor to justify an argument for purposeful evolution.
Regardless of whether I believe in evolution or intelligent design, your reasoning simply doesn’t follow.
First, adwords ads are (in the hands of a good marketer) the result of purposeful recombinations of letters and words. It doesn’t follow that mutations in viruses must also be the result of purposeful recombinations of genes.
I also don’t follow the logic of your argument for the unlikely success of random recominations of letters or genes. Many recombinations fail and quickly disappear.
The ones that live long to reproduce are the ones that get passed along.
The analogy you used is appealing but imperfect on many levels.
I think people believe whatever they want to believe emotionally. I don’t know why they (and you) can’t be content to call it faith.
If you’re going to try to justify unprovable beliefs by logic, then the logic you use ought to be tighter.
Cheers,
Dave
I have been arguing for decades that such a mechanism would evolve. Isaac Asimov postulated that some very smll organisms that have more DNA than “higher” forms of life could be using their extra DNA to contain info on how to evolve. It is one way those organisms could more quickly adapt to new conditions, and back agian when conditions change back.
perry. One of ur best posts ever. N I’ve been reading for a couple of years now. I really think u writing regular blog posts has improved ur writing. Using an adwords analogy: you cram $10 ideas/keywords into 10 cent sentences. That muliple of 1000 is were excellence resides. I love ur ability to deelve deep into an issue to find a universal truth that applies to much more than just adwords. A new standard is set. I don’t envy ur task to improve on this post…and u will…n I’ll be reading every word along the way.
Thanks for the added info. I’ll give that program and the research article a look.
I’d go a step further, and say that not only is evolution “goal-seeking” but that the entire cosmos has a “telos”, which has been given to it.
Perry, you are like an onion :-) Who knew you had so many interesting layers. Great article. Not many wordsmiths could have tied those unrelated topics into a completely understandable and relevant post.
I think the swine flu is a better lesson in how to use persuasion to influence people to do things they would otherwise never do.
Last week the news, WHO and the CDC were screaming world pandemic. Schools were closing all over the nations. Sporting events were canceled. The government paid billions to drug companies for tamiflu and relenza. A “life-saving” vaccine was started that would be proven safe by October at the earliest.
And where are we 7-10 days later?
The CDC says kids can go back to school, since the virus has spread all over the world anyways. As tragic as it is, only a handful of deaths have occurred. (A mere fraction of those killed from the “normal” flu during that same time period.)
I’m no conspiracy theorists, but anyone can see who benefited from all the scare tactics. The drug companies obviously. The government because people turned to them as their savior. And the media because more people watched the news for school closings and outbreak stats.
These guys are master marketers. To think they’ve convinced us the world will come to an end at least 3 times in the past 8 years…SARS, bird flu and now the swine flu. And they even have a continuity plan in place because there’s always the threat a new virus will mutate and emerge to cause a pandemic!
Andy,
I would encourage you to play with the Random Mutation generator. You can set it to do one mutation at a time, or 5 or 10. You will see that the more mutations you add, the faster your message gets destroyed.
What it shows you – and Shapiro’s brilliant paper explains the same thing in bio language – that a “mutator” that actually works would need to intelligently substitute verbs for verbs, phrases for phrases, adjectives for adjectives, nouns for nouns, and obey all the rules of the English language.
Shapiro is saying: THAT is how evolution does in fact work, by making very intelligent substitutions. He demonstrates it with numerous examples. This new version of evolution, this “third way” as he calls it, is LIGHT YEARS away from the traditional Darwinian doctrine. (How ironic! Evolution itself becomes proof of an even grander Design.)
I agree with you, I think that some version of evolution does match what we see in the world. But the larger point is, evolution itself is always driven by a goal-seeking program. It never, ever happens by sheer accident. As everyone who writes and re-writes Google ads proves through their daily routine.
Great post Perry. Who would thought anyone could relate the swine flu (and viral mutations) to our adwords (split testing) business! ;) Brilliant!
Tav.
Heaven forbid!
Are you saying that there could be something called “Intelligent Design”! The idea that comprehensive systems can be developed to be successful. That random mutation won’t yield success?
How about we up your bet, and wait for a random mutation to occur in Richard Dawkins, and the other “G-d Delusion” tribes. Wake up and smell the roses. The universe is not a random accident, perhaps its a series of A-B tests leading up to the grand conversion!
If an omniscient being wants to play hide & seek, and not be found. Richard and crew are certainly not going to find him. A great AD doesn’t scream AD. It prompts action.
Our universe is so well designed it screams creator & Plan, but to only those who bother to look. How many campaigns are just a piece of much more complex strategy, not just an accident.
Food for thought.
Stan
Awesome article, Perry. I love the way you weave narrative, scientific theory, marketing, theology and economics into one single blog post. From any other author, so many disparate ideas would be confusing, but you leave me totally inspired… thanks…
Thanks Perry, I’ll definitely check that out.
I’m an evangelical Christian. But I also take evolution as the most scientifically robust explanation we have for the observed world, so I’m constantly on the alert for people doing work in those areas.
Off the bat, your post raises a couple of questions. For instance, without having looked at it, does your mutation generator perform only one iteration at a time? If so, then it may be that one could improve one’s Google ads, assuming one was working on geological time scales of millions of years and had several trillions of dollars to spend on adwords. Wouldn’t that more closely resemble “evolution” as we find it in the world?
(It might not be practical to try and collect your bet, though ;-)
But anyway, thought provoking as always.
Another book you may want to check out if you haven’t already is “Finding Darwin’s God” by Brown Biology Professor Kenneth Miller.
Thanks,
Andy