Comment by koheripbal | Hacker News Reader

Comment by koheripbal | Hacker News Reader

Parent Full thread

That's the problem with evolution. Any protein in the body ends up with multiple completely unrelated functions.

If it's in the blood stream for enough generations, the body will find some alternate use for it. It's like a compression algorithm on your DNA.

...and so when we try to use meds to counter certain proteins or block receptors, we discover that it has a dozen totally unrelated impacts. "Oh, that blood pressure medication? Well let's just have it also randomly give you prostate cancer!" (real example).

Organic evolved animals are the worst type of spaghetti code - if there was ever something you'd want to redesign from the ground up, it's us.

Then add to that that there isn't even any "standard" type of human. We're all different - and even different groups of us have different smatterings of genes. Sometimes it correlates with our ancestors background - sometimes it doesn't. Who knows! Evolution! We're all each just individual experiments running around to see which genes can procreate the most.

...and it's not even as though each of us is a single experiment on a specific gene, that would be too easy. We're each an experiment on a unique collection of thousands of genes that evolution hopes will average out the results across the entire population. Try to wrap your statistics PhD around that!

Then also factor in that modern human evolutionary pressures are totally different than they were even a couple of generations ago, and it becomes clearly evident that evolution is in a complete state of dysfunctional disarray.

Spaghetti code has more negative connotation than I’d attribute to our insanely adaptive, flexible, and efficient organism.

Reminiscent of heavily optimized code perhaps?

Well we’re constructed the same way that is: by an intern that is just copy pasting from examples he’s seen work plus occasionally typing it out himself and making an error.

Except no. The error rate is very very low, and is basically the perfectly adapted “error” rate for best survival. So... not even an error.

Again, people love teaching for analogy here but humans are just nothing like code, to be honest.

Maybe the lesson is that interns are very good at copying.

The lesson is that there's millions of copies and the one that performs best on an endurance test gets to be released.

Except there's no StackOverflow to help with evolution...

Well, the difference is that for programming you almost always have an enormous surplus of computing power, while evolution is a race to the bottom on who can do most with fewer resources.

“Spaghetti code“ doesn’t imply “buggy”, so I think is a decent description. Evolution found a good local optimum, but I think that the optimal design would be quite a bit better than what we have now.

A classical example is https://en.wikipedia.org/wiki/Recurrent_laryngeal_nerve. Nobody in his right mind would design a nerve running from your brain to your larynx that way.

Spaghetti code absolutely implies buggy because code != biology. They operate completely differently, in fact potentially opposites (logic is the opposite/complement of intuition, ie neural nets, of which our body operates far more like an adaptive/analog machine than a logical one which is brutally unforgiving).

In code, a single typo breaks the whole thing. So spaghetti code by definition is bad and breaks easily.

Sure, we have a fuck ton of complexity. But it’s only messy to us because we don’t understand all the conditions that led to it.

Humans certainly have flaws, but in general we are as well suited to surviving on Earth as you possibly can be, especially on the whole.

I’d say: we’re so so much better designed than any code that exists anywhere today, that it’s a farce. We are literally the opposite of spaghetti code. We’re Antifragile!

Code cockroaches?

> In code, a single typo breaks the whole thing.

A single typo sometimes breaks the whole thing. Sometimes, it does not. The same is true for genetic mutations in humans - most typos are harmless, a few give us nasty cancers.

Yes, but that process is also what makes us live in the first place (variability). It definitely breaks, but on the majority it works better than anything we know of.

"The spaghetti code works fairly well" is a different claim than "it's not spaghetti code", though.

"Spaghetti code" doesn't mean "broken". It means "hard to maintain". Our modern medical system is pretty decent proof that maintaining a human is quite a bit of work.

Memory leaks, data loss, intermittent bugs, bad edge case handling are all examples of bugs that could result from a typo and not break the whole system outright.

Furthermore, there are plenty of examples of single gene mutations/interventions that are deadly/crippling.

Spaghetti code is spaghetti code. It refers to to the code being essentially so interwoven with each other that changing a single thing has consequences for many other locations in the code. Spaghetti code can be 100% bug free.

There are myriad of examples of our body having feature which you could equate to biological spaghetti.

> In code, a single typo breaks the whole thing.

Only in high-level languages, when you have to deal with grammar.

DNA is better compared to binary code fed to a CPU. There is no equivalent of "a single typo breaks the whole thing" there; whatever you feed the CPU, it will execute something. Whether you'll like the results or not is another question - but the same is true with biology. The DNA is there, physically, and it will interact with proteins in one way or another.

> “Spaghetti code“ doesn’t imply “buggy”, so I think is a decent description. Evolution found a good local optimum, but I think that the optimal design would be quite a bit better than what we have now.

> A classical example is https://en.wikipedia.org/wiki/Recurrent_laryngeal_nerve. Nobody in his right mind would design a nerve running from your brain to your larynx that way.

The matter is, the possibilities of DNA are infinite, you can create Godzilla if you knew how. The problem is, nobody knows how, so evolution is all we have - random mutations and survival will lead to /something/. Even genetical engineering now is in essence copy pasting of certain parts that we know work in a specific way, bound by what currently exists. But if you could really grasp biology, that works down to the single molecule and atomic levels, and actually dictate it however you like, I think it would be the final technology we would ever need.

The problem with spaghetti code is that decently intelligent people are forced to work with it and they would be vastly more productive if it weren't spaghetti.

No such advantage with our bodies, as they are only "worked on" by probabilistic processes.

> No such advantage with our bodies, as they are only "worked on" by probabilistic processes.

Well, and doctors. Who often have to deal with unintended or undesired side effects of medications and surgical procedures.

Evolution is romanticized, but its ability are limited, not infinite.

The more complex a species, the less adaptivity it has to fundamental shifts in environmental pressures. The interconnections between protein functions are directly responsible for that trade-off, and become a liability.

The time scales for evolution far beyond our current level of genetic complexity, probably extend beyond the life span of the Earth.

Now the benefit of this speghetti-code situation is that no intelligent designer is required - obviously a fundamental ingredient of abiogenesis, but does it benefit an species that's capable of genetic editing?

Nope. At this point, it's really just a bottleneck to even repairing the most straight-forward genetic diseases.

As a species, we're approaching the time where we've obviated evolution. ...I won't say we've seen that yet because reproduction rates are still very uneven across population segments, but it's only a matter of a few more generations (cataclysms not withstanding).

It’s not spaghetti code... I just think your analogy is weak and misleading.

Not sure how you’re defining complex but humans pretty soundly refute that adaptability thing.

It’s not a “bottleneck” it’s just... life. Can you point to an animal anywhere near as successful as humans that doesn’t have.. genetic variability? No. It’s a requirement and precondition for our success, not some flaw we overcame despite.

What's your definition of success?

To exist.

Existence is table stakes. All interesting things happen on top of it.

Nonetheless, it's a very freeing view to consider that just by existing you've already succeeded. Everything that comes after is an adventure. You've already beaten far, far greater odds just by being here than anything you'll accomplish in life.

is there a greater mystery than existence itself?

Something being more mysterious doesn't make it more interesting.

What's my private key? It's a mystery, but not a very interesting one.

Why does the universe exist? It's mysterious, maybe the most mysterious thing of all, and 8lit's interesting, but I'd say it's not the most interesting question of them all.

by mystery, i mean much more than that which is not known.

i mean to convey: something that is difficult or impossible to explain. something that is deeply interesting and provokes wonderment.

what is more difficult to explain and understand than the fundamental nature of existence?

perhaps it is not apparently immediately _useful_.

but in a cosmic sense, of what use is it to understand thought objects in the world when there is a complete lack of understanding of the consciousness in which these thought objects arise?

i’m curious what are some candidates for “most interesting question of them all” in your view. this is something which sparks a genuine curiosity in me and would love to hear your perspective.

Existence is a mystery indeed, but in this context, it's not an achievement. It's a starting condition for all achievements.

how do we know that achievement is of value?

how is it possible to know what is _actually_ of value as opposed to what is _apparently_ of value?

in your profile you mention:

> I do software nowadays, but I'm constantly looking for opportunities to do something actually useful for humanity.

how does one arrive at a belief about what is useful?

> how is it possible to know what is _actually_ of value as opposed to what is _apparently_ of value?

It's in our heads, but since we all run the same hardware architecture, it generalizes quite well.

it is evident that people don’t actually converge on singular sets of beliefs in this way. but for sake of argument let’s put that aside and suppose we actually do.

even with this, this doesn’t address the concern of actuality versus appearance.

consensus reality has been wrong in the past, and can be shown wrong again.

so how can you be so sure?

I think what he means by spaghetti code is that it's not readable or easy to understand or maintain by humans.

Every protein has multiple functions, so we can't make changes easily without side effects. Bottlenecking human intervention specifically.

>As a species, we're approaching the time where we've obviated evolution.

This is hubris. Everything that humanity may accomplish in the future, from editing its own genome to creating general artificial intelligence, is a product of that same evolutionary process. We do all that out of our innate need for self preservation and reproduction—evolutionary imperatives for all organisms.

But it’s more accurate to think of evolution as a universal computational process occurring within the substrate of (bio)chemistry. That computation will yield further computation (intelligence and its products).

This way you can define evolution to mean absolutely anything, at which point the word loses its meaning.

If we stick to traditional, useful definition of biological evolution, then we can confidently say that it stopped being the driving force for plant and animal life no later than when we've first learned to communicate and remember ideas; we've made it irrelevant when we've learned to write. It still works on microscale, but that's mostly because things breed and die faster at that scale, there's more of it, and we can't efficiently poke at it with any sort of precision.

Eh, you could argue evolution is still relevant in human populations up till today.

Tell that to your immune system. It's necessary, but reasonably likely to kill you or at least make your life miserable.

If we're so insanely flexible and efficient, why do we have the same body plan as every other tetrapod? We walk upright, but our weight is cantilevered out from a spine that is manifestly not designed for that.

Sure, every living creature and really any complex system has these problems: they can’t be perfect.

But you also can’t point to anything better.

> But you also can’t point to anything better.

I mean, not yet. The days of "let's tweak it so the blood vessels go behind the retina" or "let's add a bit more cartilage to the knees" aren't here yet, but they're probably coming eventually.

Evolution doesn't really optimize for the "keep this body running at age 100" scenarios that we find ourselves in lately, and there's a lot of legacy code.

Anyway, I agree there’s ways we can improve, only disagreed with the implication that we’re running on some sort of hacked together or sloppy system. The redundancy of biology is incredible, the opposite of easy to break.

Evolution is dumb. What dictates evolution is who survives. People wish to live 100 years old and be like 20, but that would be very complex, and eventually, humans exist as a species just fine without it, that is, the DNA keeps existing and replicating without that advantage. If conditions dictated that the humans who are young for 100 years leave more of their DNA than those who do not, then eventually that is the path we would have taken - maybe that is the path we WILL take, if conditions allow. But it's just a selfish wish. It's like saying an ant worker should live decades. Good for itself, maybe, but adds little value to the survival of the species, the job gets done as it is.

You are assuming evolution leads to the best adaptation. That's simply not true. It finds a local minimum and that's it. It's evident that a human that stays fertile and young for 100 years leaves more offspring then the current human. But the way to get there might not be continous improvement of fitness and thus evolution will never get there by itself.

It does not "lead" anywhere. It's a wandering in the wild. Some paths lead to cliffs, some will keep you going for a long time. A vast number of paths will be taken, but in any given time you only see the species - better yet, the individuals - who have not fallen off a cliff.

I like the idea of evolution, because "it is what it is". It happens whether you believe it or not. It has no goal, it just happens. And, it's not that evolution "changes" species, it's that whoever survives becomes the definition of what a species is at all, therefore "evolution" is just the observable outcome, the process is natural selection. So, it's pointless to "blame" evolution, to say "it is not good enough". I, you, we don't even know what's "Good enough" and for whom. Even whatever we consider "humanity" is the outcome of our natural selection. Your very wish to live forever, is the outcome of our natural selection.

The wish to genetically engineer our species' future members, will only be judged good or bad by its outcome, not by our planning. What you see now is what works, and the proof it works is we are still here. For example, if we end up destroying the world and our species with nuclear weapons, the path of evolution giving us intellect will have proven to be a bad one for this branch of DNA called "human species".

I can only say I have a very bad gut feeling about conscious genetic altering of humans, and I trust there are many arguments to find if I analyze it. Perhaps it's only some aspect of my DNA's self preservation not wanting to lose over some made-up DNA.

I donno, I am inclined to agree with the spaghetti code connotation. Humans are insanely adaptive, flexible, and efficient, but there are some serious design flaws. Here are some examples of a few things I could see improving: The need to inhale and exhale out of the same place that you need to eat and drink is bad engineering, too much risk of choking or drowning. The sexual reproductive organs sharing the space with waste elimination is unsanitary during copulation. Finally, bipedal upright legs. Legs are awkward and unbalanced, and the knees are extremely susceptible to injury, we would greatly benefit from a more robust mobility design.

Extra Credit: A prehensile tail would be extremely useful in many situations and I believe may decrease the risk of injury for elderly people. With an adequately strong prehensile tail, older people would be able to use it as a sort of a tripod leg to keep themselves upright with less risk of falling.

I'm with you on the tails. Tails would be cool.

But, I think the double use of orifices (air, food, urine, semen) is a good idea. Minimizing orifices seems to be an advantage... fewer avenues for outside stuff getting in. (And I don't think human sexual organs are necessarily dirtier due their location; urine is more or less sterile, and feces shouldn't really come into play under ordinary circumstances... )

> if there was ever something you'd want to redesign from the ground up, it's us.

Well god wrote all the code in like 7 days for this dump, so what did you expect? Quality?

Though I sometimes equate end times with "I'll get around to fixing that bug one day" as every supposed date of end times never comes. god really is a programmer.

Two widely deployed things were created over 7 days. Life, and JavaScript. Both turned out to be a huge mess.

Reply on news.ycombinator.com