Artificial DNA Replicates and 'Evolves' 126
ananyo writes "Scientists have demonstrated that several lab-made variants of DNA can store and transmit information much like the genuine article. DNA is made up of nucleic acid bases — labelled A, C, G and T — on a backbone made of phosphates and the sugar deoxyribose. The artificial polymers, dubbed XNAs, carry the normal genetic 'alphabet' on a backbone made using different sugars. The researchers engineered enzymes that transcribed DNA into the various XNAs, then back into new DNA strands. Faithful genetic transmission over successive DNA-to-XNA cycles allowed researchers to select for only those XNAs that attached to certain target proteins from a pool of random samples — a process akin to evolution over multiple generations (abstract). The research confirms for the first time that replication, heredity and evolution can take place in artificial DNA-like molecules."
Fear Not (Score:4, Insightful)
With this kind of DNA replication theory tested and proved, we now can use this to facilitate a computer chip that improves itself.
What is the learning environment that defines what "improves" means?
I've studied artificial intelligence extensively and there's a whole lot of effort going into reinforcement learning, genetic annealing, etc. But the key thing to remember with DNA is that the Earth provided this environment for it to be tested in and be given feedback. The feedback was and still is exceptionally harsh in that you either died or adapted. On top of that, the DNA lead to things that eventually competed with each other.
The problem with computer chips is that there is no fundamental death/life reward system unless we as humans implement it. And there will always be a need for us to do this because nature doesn't care about logic gates, we do. If you make a set of chips to provide an environment for incubating and reward or punishing the first set of chips, you merely have another layer where humans must evaluate and instruct the chips as to what it is that we want.
Faithful genetic transmission over successive DNA-to-XNA cycles allowed researchers to select for only those XNAs that ...
Unfortunately, in order to impose your will (no matter how lazy you wish to be) you still must define your will. And I think you'll see that it becomes a major effort when trying to set up automated systems like you propose.
Simply said: define "improves." Putting the chips outside and pitting them against each other in nature isn't gonna do it. The ability to direct harmful radiation will probably win out over gigahertz or logic gates per area.
Re:more like intelligent design than evolution... (Score:4, Insightful)
It sounds overreaching to call this "evolution" if the researchers are selecting...
Artificial selection and natural selection are equally valid ways of inducing evolution. Without some selection process, it's just random variation, with no trend.
Re:Info library for the ages stored in organisms? (Score:4, Insightful)
You want to provide proof of spontaneous self-replicating protein generation? Go find some XNA laying around outside.
There's no need, the Miller-Urey experiment neatly showed that protein building blocks can self-assemble in nature. The leap to self-replication isn't as far a leap as that between nothing and amino acids.
The fact that these people have created a synthetic analogue to DNA does not prove that a creator diety created the real stuff any more than a Tesla coil making long electrical arcs would lead one to believe that Thor throws every natural lightning bolt.
For my two penneth worth, both you and the parent are wrong, though frankly you're the more wrong of the two. kurzweilfreak is correct for the wrong reasons while you are wrong, but for equally wrong reasons. In essence, the parent is saying that a creator deity need not exist because humans can create X; you are saying that because humans made X there must be a creator. There's nothing to say a creator and human ingenuity in recreating His/Her/Its creations are mutually exclusive, and there's nothing to say that because certain complex creations have not been seen to arise in nature in human timescales they couldn't without divine intervention.
Still, I'm talking out of my arse; It's not like a have a doctorate in biochemistry (I really don't). I just get irritated by people taking interesting but relatively small - as compared to, say, DNA itself - discoveries and using them to "prove" the existence of $GOD. Personally I don't see a problem in using a deity to explain how the Universe in all its splendour was set in motion; I confess that the idea of a cosmic finger flicking the first atoms like dominoes appeals to my love of whimsy. I do, however, say that you can't reconcile evolution with a deity creating - hence creationism the Earth and all creatures thereon in their current form, since there's a more than ample fossil record that says otherwise. I'm not going to get into the cop-out that is ID, mainly because it's just that - a cop-out for school boards - but also because I have difficulty in believing in any "intelligent" designer that would reinvent the wheel. [wikipedia.org]