News for hackers. Stuff that matters.
Re:Memories of positive associations get written onto DNA by nollidge on Tuesday August 27, 2013 @11:52AM (#6284195)
I don't see why it's a paradox. Lots of things would offer fitness - why can't all animals talk? Why can't plants move? Surely these things would offer greater Darwinian fitness.
Re:Memories of positive associations get written onto DNA by samatman on Tuesday August 27, 2013 @12:10PM (#6284308)
I did want to address one other total misconception here: Darwinian fitness is a result, not an effect. There is no adaptation, at all, which predictably increases fitness. Evolution is not a direction.

"Surely these things would offer greater Darwinian fitness" is as basic a mistake as "Surely we can figure out whether this program will halt without running it".

EDIT: In fact, as I think about it, there is a deep connection between the thought "Knowing the genes of this organism, we may predict its fitness" and "Knowing the instructions of this program, we may predict its halting". I'd love to see a paper on that.

Re:Memories of positive associations get written onto DNA by tel on Tuesday August 27, 2013 @12:28PM (#6284431)
It's an effect on the side of survival, correct? It's not that some particular change was "selected" in order to increase fitness but instead that all minute changes were selected simultaneously and some were pruned. There needs to be a smooth path from the "current location in design space" to any improvement under which there is never anything damaging enough to prune off that path.

And then, this "smooth path" idea relates neatly with topological notions of computability.

Re:Memories of positive associations get written onto DNA by rflrob on Tuesday August 27, 2013 @12:30PM (#6284447)
That doesn't seem rit to me. There's lots of adaptations that could probably increase fitness of offspring as a function of the parent's environment. In the same way, you can look at some programs and predict whether they halt, just not all programs, and potentially not even all "interesting" programs, for some definition of interesting.
Re:Memories of positive associations get written onto DNA by XorNot on Tuesday August 27, 2013 @01:06PM (#6284688)
You contradict your own point though: yes, we can predict - but we can't know that a program - an arbitrary program - will halt.

Evolution is much the same, but most people are worse at it because they make the mistake of prioritizing human-values as evolutionary advantages for all species.

For example, cows get slaughtered by the billion by humans. But they're an evolutionary success story from a Darwinian perspective: by being easy to domesticate by early tribes and delicious, they've proliferated far beyond anything which would've happened un-aided by humans.

Re:Memories of positive associations get written onto DNA by dllthomas on Tuesday August 27, 2013 @01:46PM (#6284981)
For any arbitrary program, we may be able to know that it will halt.
Re:Memories of positive associations get written onto DNA by HCIdivision17 on Tuesday August 27, 2013 @05:38PM (#6286468)
For any very simple arbitrary program, perhaps. But certainly not in general. That's sort of the point of the Halting Problem: you can't know because it is undecidable.

Of course, there are simple and trivial programs that can be specifically analyzed, but we're definitely not to that level on any of the stuff under discussion here. The analogy really is pretty apt.

Re:Memories of positive associations get written onto DNA by dllthomas on Tuesday August 27, 2013 @05:56PM (#6286555)
The halting problem is undecidable much like logic is incomplete. There will always be true statements we can't prove, and there will always be programs that might or might not halt. That doesn't mean that we can't prove lots of interesting things, it just means we can't promise to prove it when handed a program. We can try, and sometimes we'll succeed.
Re:Memories of positive associations get written onto DNA by XorNot on Wednesday August 28, 2013 @06:38AM (#6288997)
The thing is a very simple program can be argued to be have been executed by the person's reading it. You know it halts because you look at the code, step through it mentally and - hey, it halts.

But that's the nature of the problem: you know it halts because just executed it till it halts.

Re:Memories of positive associations get written onto DNA by dllthomas on Wednesday August 28, 2013 @09:41AM (#6289842)
Yes, running them until they halt is one way of determining that a program halts on a given input. However, we can prove things do or don't halt without that. Consider running, on an idealized machine (so no overflows),

    while(i>0) { i * 2; }
I think you would agree that this clearly never halts for any i > 0, and I think it also clear that we did not determine that by running it until it did not halt (because there is no such point).

Turing proved that for some programs there is nothing we can do but run them and see. That doesn't mean that's all we can ever do.

Re:Memories of positive associations get written onto DNA by nollidge on Tuesday August 27, 2013 @02:07PM (#6285097)
Yep, that's more or less what I was driving at, but I was being a bit more coy.
Re:Memories of positive associations get written onto DNA by kbenson on Tuesday August 27, 2013 @03:18PM (#6285547)
You mean Darwin's Non-halting theorem? ;)
Re:Memories of positive associations get written onto DNA by samatman on Tuesday August 27, 2013 @12:00PM (#6284257)
It's not a paradox, it's an almost paradox, which is kind of like being almost pregnant from a logic perspective.

Lamarckian and Darwinian evolution were debated and Darwin won. Lamarck's Lemma, which is a term I made up, points out that, were a random walk to create a mechanism for Lamarckian evolution, Darwinian evolution might well keep it around, as it could offer fitness.

That's why I phrase it as a lemma on Darwin's theory: even if it proves true, as epigenetics suggests, it doesn't make Lamarckian thinking correct. If anything, the opposite: it shows that Lamarckian effects can be explained through natural selection.

Plants can of course move, the classic example being the sensitive plant. Whether or not non-human animals can "talk" depends, basically, on whether one means "signalling" or "language proper". The bee dance can give arbitrarily complex instructions in one particular domain: there is no 'one' bee dance, each performance is roughly as unique as a sentence or so of text.

Re:Memories of positive associations get written onto DNA by VladRussian2 on Tuesday August 27, 2013 @03:45PM (#6285765)
> Whether or not non-human animals can "talk" depends,

there is at least one animal that we know is able to talk, at least in our definition of "talk" - this animal is humans, so the Evolution did produce a talking animal.

Re:Memories of positive associations get written onto DNA by jonnathanson on Tuesday August 27, 2013 @11:42PM (#6287909)
Darwinian natural selection doesn't select for absolute/universal fitness, because there is no such thing. It selects for fitness into a specific, local niche. Humans may be more fit than chimpanzees in some areas, for example, but less fit in others.

It is a fallacy to assume that evolution is linear, and that it progresses toward some sort of universal optimum. There is no universal optimum. There are only specific, local optima.

A good example is the megalodon. From all we know, the megalodon was basically a massive, faster, more badass, more ferocious version of the great white shark. Why did the megalodon die out, and the lines that led to the great white survive? Surely big-badassedness is a desirable quality in an apex predator like a shark, right? Well, not really, because circumstances changed. Ocean temperatures changed, prey got scarcer and smaller, and eventually, size became a detriment. Megalodon was too big not to fail. What fit one day no longer fit the next.

The same thing may happen to humans someday, and some other animal may step up to take our places. It seems unlikely, but nature is the one doing the "selecting" in natural selection -- not us. Environments and circumstances change, and as they change, so do the criteria that "fit" the new context. In the Darwinian sense, "fit" refers to suitability to a habitat; it does not refer to "fitness" in the modern sense of the word. Think "product/market fit," not "physically fit."

Re:Memories of positive associations get written onto DNA by sadkingbilly on Tuesday August 27, 2013 @12:47PM (#6284565)
> Why can't plants move? Surely these things would offer greater Darwinian fitness.

Not necessarily for plants. They do in fact move. They just do it differently. They have no problem spreading their DNA throughout the world via seeds and pollen. They use a lot less energy too, compared with sprouting legs and getting up to walk around. If they did this, they would likely require much greater quantities of food, including protein, which would put plants at a disadvantage.

In any case, plants have also evolved on a completely different fitness track than that of animals. Plants have evolved incredible manipulation of chemistry, far beyond what humans can produce today, let alone even understand.

I recently finished reading Botany of Desire (Michael Pollan) where he describes how plants have co-evolved with humans, appealing to our selection of flowers, fruits, and vegetables to further distribute their DNA. Who's actually farming who? Interesting stuff.

Re:Memories of positive associations get written onto DNA by nollidge on Tuesday August 27, 2013 @02:12PM (#6285125)
I was being a bit tongue-in-cheek. You're absolutely right that supposed "benefits" might have unforeseen tradeoffs. In this case, having some sort of pseudo-Lamarckian inheritance could possibly be a detriment due to energy requirements or something like that. Or indeed it could have a neutral effect on fitness, in which case it's unlikely to take hold in the population, barring genetic drift or other non-selection evolutionary mechanisms.
Re:Memories of positive associations get written onto DNA by VladRussian2 on Tuesday August 27, 2013 @03:40PM (#6285733)
>I don't see why it's a paradox. Lots of things would offer fitness - why can't all animals talk? Why can't plants move? Surely these things would offer greater Darwinian fitness.

this is why we have live organisms that able to move - it is called animals, and we have animals that can talk, and we'll possibly have [some milleniums later] animals that are able to efficiently transfer memory through DNA (possibly after these animals become able to calculate and select what memories they'd like to encode into their own DNAs and what of these memories they want to transfer to the offsprings)... Evolution isn't finished as of today, it is possibly only starting :)

Re:Memories of positive associations get written onto DNA by j2d3 on Tuesday August 27, 2013 @05:24PM (#6286395)
Animals can talk --- to each other.