Hacker News Reader

Comment by sethbannon | Hacker News Reader

I am so insanely excited for the potential of this technology. There are many ethical questions here, but the potential benefits far outweigh the downsides. In the near future, we can detect and eliminate genetic disorders, ensuring no child has to suffer from these defects any longer. Long term, this gives us a tool to take control of our own evolution in a way never before possible.

Couldn't be more excited for what's possible.

View on HN

While I appreciate the optimism in this sentiment, the similarity to the early atomic age gives me pause. Human beings have shown time and time again that they cannot be absolutely trusted.

Yes, the worst really must be assumed

The atomic age came without the world or our species ending.

There's an anthropic problem there. Unless you assume humans are alone in a single universe with no branching, or that there are a large number of worlds where an atomic age (as well as any other equivalently stage of dangerous technological advancement) is infeasible, surviving the atomic age tells us that the probability of us surviving events in that reference class is non-zero, but it doesn't tell us much more.

We are still in the "atomic age" - and will be for as long as nuclear weapons continue to exist.

Furthermore, right now these weapons just "sit around" - but one day they will be used again. All it will take will be a "brief moment" of insanity by a leader, and by the people he leads, to touch off the tragedy to follow.

Or...more likely...because we know it has happened in the past: It will happen due to a mistake, either by one or more humans misinterpreting what other humans are doing, or by misinterpreting the signals of machines (or machines sending the wrong signals, which are then interpreted correctly by the humans - though the signals are false).

This scenario is far more likely than sheer anger causing madness leading to tragedy - simply because it has historical precedent of happening not once, but multiple times.

Humanity will never be safe as long as these weapons and weapon systems exist.

By that token, every single activity of anyone, anywhere, any time has been a success.

So let's put it differently: was the atomic age net good for humanity? I'd say very much, yes.

Maybe. I think it's difficult painting these things as having a net value. There's good and bad, and they co-exist, one doesn't come without the other.

Sort of I guess, every activity puts us in a new place. We know we can't stand still. Doesn't mean we can't be careful, but handwringing won't help much.

Why vest absolute trust in anything at all? Just figure out how much trust is justified and act accordingly. And remember: the conservative, do-nothing choice is also a choice, because the world still does its thing even if you do nothing.

I agree for deadly genetic disorders, but we can't be sure if a harmful mutation will be important in the future. For example sickle-cell disease make people immune to malaria.

What if in the future a pandemic rises, and we have "edited away" the only mutation that would give immunity?

Unless you modify everyone ever, this is extremely unlikely to happen.

There are people resistant to malaria that do not carry the deleterious red blood cell mutation, but a different variant.

Why is that unlikely? As cost approaches zero and benefits increase we can assume we get pretty close to 100%. In fact holding a portion of the population back, even if well meaning, will be discriminatory.

I give us decent odds we end up like the bananas.

1) It's not perfect immunity but rather resistance, and it's when they are heterozygous.

2) Heterozygous still carry the gene and can pass it on (you're likely to just exclude homozygous people)

3) Sickle-cell disease has a lot of downsides... Yes, it's a disease.

There is little to zero downside in preventing births of homozygous affected people.

You simulate and test in lab for an even better mutation. Guided evolution is much faster than genetic evolution.

We could get rid of genetic diseases without editing the genome. Just screen embryos before IVF.

this does not edit embryos, this only select the ones without defect and trash the defective ones. Very different from editing, especially for couple that only get bad embryos, they'd love to "fix" it.

Or they could use a different source for sperm or eggs, especially that they are supposedly not attached to their own genetic code (being willing to "edit" it anyway).

> especially that they are supposedly not attached to their own genetic code (being willing to "edit" it anyway)

There's a difference between editing and replacing. Imagine my father gives me his watch (which he got from his father, etc.). The watch is fine, but the belt has wear damages and the glass could use cleaning. Doing those repairs doesn't mean the watch loses sentimental value for me, and is markedly different from just getting a new watch.

That's why he said without editing.

We can already detect and eliminate genetic disorders.

Just barely... only for single nucleotide disorders and we are finding that a lot of the coding in the genome is not continuous and can be broken up by hundreds if not thousands of base pairs.

Deep learning texhniques should help us identify the genetic causes of sone diseases but I suspect that our microbiome and epigenetic factors play a much larger role than currently known.

Have you read some dystopian fiction? For example, Margaret Atwood's Oryx and Crake?

I'm excited to see what we can do with CRISPR, but we should do it fully aware of the ways things could go wrong (so that we can prevent those things from happening).

Reply on news.ycombinator.com