Ask HN: What scientific phenomenon do you wish someone would explain better?

I'll take a shot at trying to explain that to the best of my ability. I'm starting from the point where we have organisms and dna, since nobody knows how exactly abiogenesis happened yet.

First, two general points.

The most important thing to keep in mind is time. Life has had insane amounts of time. Billions of years. Beyond human comprehension amounts of time.

The second most important thing is that complicated != carefully orchestrated or optimal. Life is pretty cool, but it doesn't hold itself to a very high standard. It's the survival of the good enough, and is full of so many random hacks and poor design choices it's insane. Things get easier to accomplish when you lower the standard.

Now an attempt at an explanation.

Evolution by natural selection works on two principles. First, generation of diversity. Second, selective pressure.

DNA can and does mutate frequently. One important type of mutation is a duplication, since it let's you gain new functionality. You make two versions of the same gene, one keeps it's original function, and the other does some new function. This theme of repurposing existing things comes up again and again. Take something you have, make another version of it, change it a bit. If you've worked out how to grow a vertebra as a lizard and want to become a snake, turn off legs amd make more vertebra. Use the same genes, and just modify how you control them. This video (https://youtu.be/ydqReeTV_vk) is actually pretty good at running through the science behind evolutionary development, and how evolution can quickly reuse and modify existing parts. Basically, once smaller features evolve, you start modifying them in a modular way and can start making really big changes really easily. Keep in mind again that you are helped by enormous, mind boggling amounts of time randomly generating these features originally as well. Heres a summary of how our eyes evolved by repurposing neurons (https://youtu.be/ygdE93SdBCY). Our eye is a good example of how the standards are just good enough, we have veins and nerves on top of our light detecting cells instead of behind and just poke a hole to get them through to the other side. Doesn't that leave a blind spot? Yep, and we just hallucinate something to fill in the space. There are a couple other major ways we generate more diversity. You have things like viruses transferring DNA, but a really powerful one is sex. Sexual reproduction lets you combine and generate new combinations of Gene's to speed up how quickly diversification happens.

For selective pressure, think about it purely statistically. You have sets of arrangements of atoms, some of which are good at making new sets of arrangements of atoms that look like then, and others less so. Each tick of the clock, versions that are able to make more increase, and versions that dont decrease. This basically provides a directionality for evolution - whatever is good at replicating is successful. This weeds out mutations that dont over time while keeping mutations that do. This means the next round of mutations are building on ones that were good enough, and not ones that weren't. This lets evolution be more cumulative than a random search.

Neither of those are complete descriptions by any stretch, I'm just trying to give you a taste of the mechanisms behind it, but it goes a lot deeper. The most important things do just boil back down to what I started with. Survival of the good enough - lower your standards. And there's just so much time for these to happen. Evolutionary step only has a 1 in a million chance to happen in a given year? Then it's happened about 65 times since dinosaurs went extinct.

A core point of relativity is reference frames. From your perspective light in a vacuum always travels at c. The problem is that by special relativity, an observer moving at some speed relative to you will also see light traveling at c. This is a simple idea, but it causes a lot of very unintuitive effects. If we're being pedantic, relativity is very logical mathematically, but it's conceptually difficult because it flies in the face of how you think the world works.

Have a look at the simple inference example here: https://en.m.wikipedia.org/wiki/Time_dilation

Time doesn't necessarily slow down the further away you get from a clock. If you and a clock are both stationary (ie you're in the same inertial frame), you will observe it ticking in "normal" time, albeit delayed due to the distance. If the clock is moving relative to you however, you will measure its ticks to be slightly slower.

You may be confusing general relativistic effects which are distance dependent (as gravity weakens the further away you get).

If you carry a clock in your rocket, you will (in the rocket) measure it to tick once a second. When you get back to Earth, you'll find that it's lagged behind a clock that was started at the same time but was left on Earth.

Maybe have a look at simple wiki too https://simple.m.wikipedia.org/wiki/Special_relativity though it doesn't actually derive the Lorentz transforms unfortunately.

Ignore the gravity bit for now, that's general relativity and it's more complicated to explain.

I have to say that video completely changed the level of my understanding about it. Especially the bit of visually intuitively understanding why the imaginary terms are the integration of the wrapping of the frequency component. Well worth watching.

Since no one has listed it yet, please check out https://quantum.country/

It's by Andy Matuschak and Michael Nielsen, and it is excellent. Have fun!

Also, for me, how all these advances in ML/computing are alleged to be on the horizon, when I hear that A/D/C(NN)s are actually so slow in learning. How can something that has to be trained with 1M of be smart? What's the next thing?