Hacker News Reader

Comment by dna_polymerase | Hacker News Reader

> “The startling regrowth of tissue in this case occurred because we were able to intervene early in a growing child, before long-term tissue degeneration. Although it’s impossible to conclude from this single case if the sequential application of normobaric oxygen then HBOT would be more effective than HBOT alone, in the absence of HBOT therapy, short duration, repetitive normobaric oxygen therapy may be an option until HBOT is available."

It was done for the first time, we have no idea what really did the magic here, we can just assume. In any future medicine scenario I'd expect us to exactly know what we are doing and how it works. Yes medicine is improving but we really are in the dark ages still.

View on HN

>we really are in the dark ages still

No, we'd be in the dark ages if we were in still the dark ages. Our current level of progress would seem like science fiction or even magic as little as a century ago.

I think I understand your point, that we have so much more to learn we're not even able to quantify what we don't know.

But this is not some kind of endless quest. We're dealing with a machine in the end with finite complexity and eventually it will be understood fundamentally.

I don't think I would be surprised if the most common diseases were eliminated in the next 100 to 200 years.

>But this is not some kind of endless quest. We're dealing with a machine in the end with finite complexity and eventually it will be understood fundamentally.

I'm guessing you don't work in the medical field (I have). First their's the emergent complexity of scale (after all the universe consists of a finite number of a small number of fundamental particles). Second, the human body (and all the pathogens etc) are object code highly optimized over billions of years. We understand almost nothing about how it all works. We can transcribe most of a human's genome but barely understand how even a tiny part of it works.

> I don't think I would be surprised if the most common diseases were eliminated in the next 100 to 200 years.

The most common causes of death have been pretty much eliminated in the wealthy world over the last 50 years. People still die, just of a wider variety of less common reasons. In the limiting case, each person will die of a unique cause.

The point is, even if the complexity is large, it's not infinite. And over the last few centuries we got pretty good at developing both mental and technological tools to tackle the complexity. While a human can't be expected to keep track of e.g. all the methabolic pathways described in medical textbooks and research papers, it's not a huge deal for a computer. Now of course that is only one of the points in a web of complexity that recursively expands in breadth and depth - but that only invites us to build more and better tools.

> People still die, just of a wider variety of less common reasons. In the limiting case, each person will die of a unique cause.

Right now, human lives are still capped at ~120 years, and usually much less than 100. In the (mathematical) limit, each person will die of a unique cause at age infinite. While we won't ever reach that limit, it's worth to continue getting closer to it.

To extend your metaphor a little, we can transcribe a pseudo source-code (nuclear DNA), a preprocessor or macro language that we have learned a little about (epigenetic regulation), and there are still two levels of IR after that. The object code might be the trillions of combinations of amino acids that make up the proteome.

>guessing you don't work in the medical field

Maybe the fact I don't actually makes it easier to understand biological complexity.

You guys have to figure out how to fix a brain when it breaks.

In CS, a holy grail is to match or exceed human cognition with general AI. It's not fixing something broken, its reverse engineering or alternatively engineering from scratch the greatest known computational device in the universe which has evolved over a billion years.

It's often not appreciated that this hasn't happened yet not due to lack of computing power. It hasn't happened yet because we're still not even remotely close to knowing what code to write if we did have 10,000 times more compute. We don't know how to do it, and instead are forced to plod along in awe as every year we learn a little more about what we don't know.

> I think I understand your point, that we have so much more to learn we're not even able to quantify what we don't know. But this is not some kind of endless quest. We're dealing with a machine in the end with finite complexity and eventually it will be understood fundamentally.

I see this as agreeing with the person you're replying to. We can make certain limited interventions which are amazing, but the kind of complete understanding that you're referring to is a long, long way away. Even when we do have a treatment, we often do not fully understand how they work.

A little over a century ago and there was still debate about germ theory.

Unfortunately there are still some people who don't believe it today :(

I think we have already done a bangup job eliminating "common diseases" over the past 70 years.

Now we just have to eliminate stupidity so all those hard-won gains don't get undone because of a handful of anti-vax types.

And coordination problems, so that if we succeed with anti-vaxxers we won't get killed by farmers abusing antibiotics on farm animals.

Nice side effect of solving coordination problems include: solving poverty, solving peace on Earth. :).

> we have no idea what really did the magic here, we can just assume

...which is true of most of medicine. Most indications (e.g. diseases) are named after their symptoms rather than causes, and a lot of treatment is just creating / improving the environment in which the body can do its job.

None of which should be taken as denigration of amazing achievements in the medical field!

Exactly. I recognise all those achievements, yet we still have to learn so much more.

Reply on news.ycombinator.com