Synthesis is a good thing: Mark Chu-Carroll finds a similarity between Notch signaling and something mysterious called The Itai-Rodeh solution. It’s not surprising that similar algorithms might arise by design and by evolution—which is why the IDists have to demonstrate something unique to their assertions for which a mechanism without a designer cannot account.
Pharyngula
Evolution, development, and random biological ejaculations from a godless liberal
Paul wrote:
“It’s not surprising that similar algorithms might arise by design and by evolution.”
The simple flaw in your thinking and in Chu-Carroll’s thinking is that ALL algorithms are the product of intelligent input. Whether they are designed, or whether they “evolved” an intelligence was required.
What you need to demonstrate is an algorithm that has emerged WITHOUT the benefit of intelligent input. Now that would be something!
Two nice posts, due the pathways of the web. And we don’t even need to find out a leader.
Two nice posts, due the pathways of the web. And we don’t even need to find out a leader.
The tone of this post makes it seem as if PZ isn’t familiar with the large body of work done on genetic algorithms – one early researcher is Dr John Koza. It is well known that a simple computational method can duplicate forms and behaviours found in nature and also methods and designs independently invented by man – and frequently outperforming them.
As for wagner’s assertion that “intelligence is required”: A simple examination of genetic algorithms proves that there is no intelligence whatsoever in such a system; merely a simple mechanical definition of reproduction and selection. A system of this sort – capable of generating complex systems – is no more intelligent than a mousetrap. In fact the very success of genetic algorithms alone is a complete experimental validation that the mechanism of selection is sufficient to explain both Natural and man-made devices.
just when you think you’ve found a nice post to have an interesting discussion, you find that charlie has already farted in the room.
can you just let us discuss biology for once?
“What you need to demonstrate is an algorithm that has emerged WITHOUT the benefit of intelligent input. Now that would be something!”
How would one do that? It seems to me that you’ve already made your assumption that all complexity and self-organisation is the result of some form of intelligent intervention, so how could one show you something that isn’t? Any biological system you will say is ID, any model you will say is a result of the human intelligence pre-loading or designing the model.
It seems clear that you simply don’t actually understand GAs or how the sheer amount of trials and generations in nature allows novelty and speciation.
Hmm, yes, my earlier post never appeared. But only a passing familiarity with genetic algorithms (e.g. work of Dr John Koza in electronic circuit design, etc) is required to see that there is no “intelligent” component. The mechanism is simple and “stupid”.
ALL algorithms are the product of intelligent input? Horsepuckey.
Physical and chemical laws provide many mathematical combinations for use in algorithms, but the algorithms need not be designed. We can design algorithms that simulate nature, but we should not presume that our formulae cause the effect in nature.
Emergent, not designed.
Chemical equilibria, hydrophilia, heat transfer, charge dispersion all contribute their statistical, yet often very predictable, math to the fray. Combining them, you can get other effects, some digital (as in voltage thresholds in neurons, or some gene activation in DNA), some more affected than others by inputs (e.g. temperature).
I didn’t know Itai was an actual name. I thought it was the manga exclamation for “ouch!” :)
Well, given that computers are specifically designed not to produce unpredictable results and as such contain absolutely no random elements there’s always going to be a hollow Get-Out-Of-Jail card for ID supporters.
Due to their nature, demonstrating evolution on a computer will obviously require creating some virtual environment that variation and selection can take place within. Then the ID supporter can always just point to the creation of the environment and say “Hey, that’s where the intelligent agency went in!”.
However, applying this idea back to the real world just pushes God back to the standard creating-the-Universe position, and evolution then still took place, so it doesn’t really help them out at all.
Charlie Wagner: “The simple flaw in your thinking and in Chu-Carroll’s thinking is that ALL algorithms are the product of intelligent input. Whether they are designed, or whether they ‘evolved’ an intelligence was required.”
A friend of mine works for the military side of Boeing. Among the things he works on are algorithms that target objects with unpredictable movements. Algorithms with “intelligent” input from people simply can’t do the job well enough. What’s used instead are “genetic” algorithms: algorithms with random, undirected changes (mutations, if you like) compete against each other, those that hit the target most often have further random, undirected changes made in them, and so on. Because computers can run a huge number of trials so rapidly, within a relatively short time algorithms emerge from this random evolutionary process that perform much, much better than anything intentionally designed by humans.
So it seems like evolved algorithms without ‘intelligent’ input during the evolutionary process work just fine (far better, in fact, than those that *are* intentionally designed) for very critical real world purposes.
Jud wrote:
“So it seems like evolved algorithms without ‘intelligent’ input during the evolutionary process work just fine…”
All algorithms, including your examples, require intelligent input to create. There are no exceptions.
The idea that these algorithms “evolved” without intelligent input is illusory. while the algorithms “target objects with unpredictable movements”, the algorithms themselves are the product of intelligent (human in this case) input. Because the computer is creating the algorithm, it creates the illusion that it is without intelligent input. But who created the computer that creates the algorithm that solves the problem?
Dr Frank wrote:
trollie wagner wrote:
So predictable.
But who created the computer that creates the algorithm that solves the problem?
By this logic it is completely impossible to ever do any kind of experiment to support evolution. Hey, even if you shoved some amino acids in a big jar and, after only a few weeks, miniature tentacled creatures that could beat you at chess had spontaneously appeared, it still wouldn’t prove evolution as an intelligent agent was behind the whole thing. Which is, of course, just a silly way of looking at things.
At best, such an approach could end up being a fine-tuning argument, but it certainly wouldn’t validate anything to do with biological ID.
The idea that these algorithms “evolved” without intelligent input is illusory. while the algorithms “target objects with unpredictable movements”, the algorithms themselves are the product of intelligent (human in this case) input.
See, I know you think it it illusory, but you use an irrelevant (straw man) argument. The algorithms mutate randomly and are selected based on there successful adaptation to their enviornment – hitting the target. The origins of the enviornment are irrelevant, FFS, the fact that genetic algorithms adapt due to random mutation is what is being tested.
Furthermore, if you have a beef with the enviornment, show that the parameters are wrong. As someone said above, a computer is necessary to perform these types of simulations because natural processes take much longer than human lifetimes, even the total time of our species existence, so it is really stupid to just create another earth to see if and how evolution takes place so we can document it. In fact, there is fossil and genetic evidence all over the place (which is of course the documentation you imply needs to be seen occur in order to produce ‘real time’ observation to document). This is the basis for the ‘random mutation leading to advanced algorithms’ hypothesis in the first place, and the computer simulation is a test of this hypothesis, whether a man made the computer, or it arose spontaneously – it is effin’ irrelevant. Every experiment to test every hypothesis involves our input, but that does not mean anything whatsoever about the original observations and their enviornment.
Just because one enviornment, the computer/lab/classroom/etc is synthetic does not mean shit about the originally observed enviornment.
This is another false analogy you dweebs like to push: if to things are alike in one way, they are alike in every way.
Any ten year old with a modicum of common sense can tell you this.
Because the computer is creating the algorithm, it creates the illusion that it is without intelligent input. But who created the computer that creates the algorithm that solves the problem?
Saying the same thing over and over, just in different wording does make it true, it just illuminates your inability to understand other considerations.
It doesn’t matter who created it.
Hey!!!!
By your reasoning, someone or something would have had to create whatever ‘intelligence’ is behind intelligent design, or god, for that matter.
This is known as the principle of first causes, and you beautifully illustrate the futility of trying to argue for intelligence creating anything complex. That would have to include god or whatever, so STFU!
Y’know, this whole “Intelligent Designer” is actually a red herring of sorts, or an unbased assumption at least.
Any claim that there must be an ‘intelligence’ behind evolution really, and I mean REALLY, cries for an explanation of where that intelligence came from, for it must have been created (due to it being too complex to have arisen spontaneously) and this is the crux of IDers arguments.
The burden of proof is obviously on them, it is a stupid idea in the first place – we all know that – but, arguing with them is futile anyway and we don’t help by giving ID credibility (in their minds) by trying to show how them how nature works randomly. Make those f***ers back up their side before we allow them to use it in an argument.
So how about it paul, where the f did this intelligence come from?
Sorry, CHARLIE, not paul
Sorry, Charlie, your red herring won’t make the cut for our delicious canned tuna.
Actually, Charlies argument is mute for an entirely different reason. No matter how good your “intelligent” input, the algorythms won’t “necessarilly” be doing just, only or exactly what you intended them to. You can tweak the system to try to *force* them to do what you want, but they can surprise you, like the algorythms that the Avida people evolved, which, over time, also evolved the ability to appear inactive when dropped into a test environment. They survived the experimentors attempts to kill them by slowly adapting changes that “recognized” the data fed to them in the test system and ignoring it. Now.. Its obvious what happened. Instead of killing off “all” evolving versions, you mistakenly killed off only those active in the test system. The next generation was already “less” active in the test system, so the ones that where even less active there (or just as inactive). The problem with “intelligent input” BS is that evolution evolves the way **it** wants. In a very precisely controlled set of conditions, like one that tracks targets, the entire environment is so limited in scope that the **only** conditions they *can* adapt to are the ones intelligently introduced. On something like a planet… This is totally impossible. Even if you created some vague “master script”, like charlie babbles about, random changes from radiation, drastic environmental shifts or just the availability of food from year to year would alter both the genetic algorythms **and** the script itself beyond all recognition in less than a million generations. The “script writters” might have intended intelligent spacefairing squid and ended up with apes, and *that* is just assuming it didn’t fail entirely at some critical step and produce nothing but giant, and sub-intelligent bugs or something.
Put simply, *we* can’t control evolution of algorythms over a span of a few years in a computer simulation without limiting the simulation to something about a billion trillion times less complex than the real world. Even something 1,000 times more complex than the military’s target tracking system we lose control of. Yet, Charlie would have us believe that some species, from some place, popped in billions of years ago, dumped a genetic algorythm on the planet and that the “result” is exactly as they intended it…. You might as well say God, for how ludicrous such a suggestion actually is.
See, this is what’s really funny though. Taken logically, his arguments don’t even make sense. In his version, there is some “predetermined” path that everything goes through to get to us… That means that mutation **doesn’t** happen at all, its all pre-programmed, right down to the last tiny change that happens at every step. Even if its only “generally” guided, its **still** being specifically guided to a specific result. Only problem is, that require “something” to fine tune things all the way through the entire process. Most projects on evolution simulations, like Avida ***never*** fine tune. There is now “master” script to follow, no “goal”, no “final result that is aimed for”. Once someone has tossed the basic critter in the woodpile, everyone stands back and just watches what it turns into. On a basic and fundamental level, non-directed, non-applied genetic algorythms, which only provide environment and no “direction” other than “survive” resemble Charlies version of evolution about as much as a twig resembles a space shuttle.
I say: “It seems to me that you’ve already made your assumption that all complexity and self-organisation is the result of some form of intelligent intervention, so how could one show you something that isn’t?”
Wagner says “The idea that these algorithms “evolved” without intelligent input is illusory.”
To paraphrase House MD, I say it here and it comes out there.
Gah! Why is Charlie Wagner still around? I thought you were getting rid of him finally. His posts take up space and haven’t contributed anything interesting or new in a very very long time.