Hacker News Reader

Comment by duxup | Hacker News Reader

A recent PBS documentary did a great job talking about all the changes the earth has gone through:

https://www.pbs.org/wgbh/nova/article/the-history-of-earth-i...

They had a series on planets and some of their history too.

Growing up I honestly thought that planets evolved into a sort of natural stable state that I guess I assumed would last until the sun or other external force decided otherwise. But it’s interesting how throughout time the earth has dramatically changed due to “minor” changes like volcanic activity changes, the introduction of, water, oxygen, or bacteria or other chemicals.

I had thought the sun and distance from it really was the determining factor as far as the temperature generally goes.

View on HN

Well, it's a really good start. Total insolation is a great, ah, trend marker. Higher insolation planets tend to be warmer.

Of course, insolation itself stops looking less like a number and more like a function. After all, you're getting photons of different wavelengths (pretend the sun in question is a blackbody) and so that feeds into absorption, reflection, and transmission of each frequency component versus the local atmosphere.

Local atmosphere is a function of surface gravity, orbital velocity (you have to scoop up those gases after all), geologic activity, possible tidal lock, and temperature. Whoops! Back to where we started.

But it is still a great trend. Very low-mass planets are in some ways easier to model. If they are far from a star, any gases that wouldn't escape would freeze and just fall down as another layer of the planet. If they are close to a star, you're just not going to have much atmosphere for long.

Once you are big enough to retain an atmosphere, stuff can get wonky.

> Growing up I honestly thought that planets evolved into a sort of natural stable state that I guess I assumed would last

This seems to be another expression of the truth that "people tend to mistake the limits of their view for the limits of the world."

Rumsfeld’s unknown unknowns.

Rumsfeld is famous for the quip, but it's basically an expression of the first stage of the four stages of competence:

1. You don't know what you don't know.

2. You know what you don't know.

3. You know what you know.

4. You don't know what you know.

https://en.wikipedia.org/wiki/Four_stages_of_competence

Rumsfeld was referring to a way of categorizing uncertainties in a mutually exclusive and collectively exhaustive way. There are many scenarios that are genuinely unknowable - it has nothing to do with competence. He probably saw the matrix on a McKinsey consultants slide earlier that day while they were using their 6 months of strategy consulting experience out of college to help him run the war.

No, it is much closer to the Cynefin framework's understanding that complex systems are affected by the influence of variables which are too small for us to measure or even know they exist until after they act.

Rumsfeld's embarrassment over that line was well earned. He was trying to sell a war based on Iraq having weapons of mass destruction. He was asked why we should go along when there was no evidence for that. He tried to sound smart saying, "Sometimes you don't know what you don't know", basically admitting that there was no evidence to justify his predetermined course of action.

They weren't "unknown unknowns". We knew. He knew too, but he was trying to muddy the waters. He got his way at the expense of one million dead Iraqi civilians, thousands of dead US soldiers, and 3 trillion dollars of US money.

The Rumsfeld formulation itself dates to at least the 1950s and development of the Johari window (named for Joseph Luft and Harrington Ingham).

I'd first encountered it in print in a 1982 book, Forced Options, by Roger Lincoln Shinn, two decades before Rumsfeld's famous 2002 usage:

<https://archive.org/details/forcedoptionssoc0000shin_c8x7/pa...>

Never heard of it.

Wittgenstein reference no?

"Pardon him, Theodotus. He is a barbarian, and thinks that the customs of his tribe and island are the laws of nature."

- George Bernard Shaw, "Ceasar and Cleopatra"

I wonder if we’ll get to a computational space where we can completely simulate every atom in a small closed system. Are we there yet? Can you simulate at the atomic scale and get a perfect representation of reality or do we have to try and simulate something smaller?

It's a bit of a stretch perfectly simulating one proton, I'm afraid.

https://www.quantamagazine.org/inside-the-proton-the-most-co...

Oh wow. Maybe that can be a new scale measure of technological capability based on how much of reality we can perfectly simulate.

If you think about it, a computer's not going to be able to perfectly simulate it's own physical operation, so the limit of a system's ability to simulate things is it's own complexity.

This isn’t obvious to me. It doesn’t have to simulate itself in real time. It may not be able to simulate itself simulating itself. There may be proof of this using a diagonalization proof.

Consider programs that are quines, programs are able to output the exact source code of the program. See [1].

And there are abstract computers that can produce any computable output. These are called Universal Turing Machines, see [2]. UTMs can be specified with a remarkably small number of internal states, see [3].

I’m not saying you are wrong, but computability is full of unintuitive results, and the answer may be more subtle than what is revealed by “just thinking about it”.

[1] https://en.m.wikipedia.org/wiki/Quine_(computing)

[2] https://en.m.wikipedia.org/wiki/Universal_Turing_machine

[3] https://en.m.wikipedia.org/wiki/Wolfram%27s_2-state_3-symbol...

It needs to be able to store a complete representation of different versions of it's own state, in it's own state.

We can get up to 100 or so lightweight atoms (the real limiter is the number of electrons, not the number of atoms. Larger atoms cost more). Maybe a thousand or so on supercomputers.

Quantum computing will change this though.

Is it an absolutely perfect simulation? Like would we know the exact future of those hundreds of atoms in perpetuity?

No, for multiple reasons. To your first question, there's still simplifying assumptions. These things usually treat the nucleus as fixed and static, for example, and often other approximating assumptions are used. Outside of the interior of a star, radioactive elements, or when relativistic energies are involved, these are good approximations though.

More fundamentally, quantum systems are inherently random. You literally can't know the exact future of hundreds of atoms in perpetuity--that's not how quantum physics works.

But you can get an idea of minimum energy states, the energy barriers holding those states in place, and therefore what the stable outcomes might be.

The other reason the answer to this is "no" is that we don't even really understand how an atom even works its entirety yet. We don't know fundamentally know what makes the potential energy of an atomic nuclear be structured in the way it is.

EDIT: I thought I'd double check that since it's been a while since I studied physics formally; apparently it is possible to model light nuclei[1] (up to four nucleons) from the ground up now!

[1] https://en.m.wikipedia.org/wiki/Ab_initio_methods_(nuclear_p...

[1]

You are confusing exact solutions with the ability to solve. We know how atoms work and can solve for any property we might want to measure. Ab initio systems aren't very practical for large systems because of computational costs, but they would work if given infinite compute (or a quantum computer). We just solve them numerically instead of closed-form.

It's like polynomials: there's a quadratic formula, and a cubic and even quartic formula for exact solutions (although the quartic is ridiculous and no one would ever use it). Higher order polynomials have NO exact, closed-form solutions using standard arithmetic operations. But we can still solve any polynomial no matter the degree using numerical methods. To say that we don't even know how how degree 5 polynomials work would be flat out wrong.

I’d recommend reading up on Heisenberg’s Uncertainty Principle to understand why what you suggest is, to our knowledge, physically impossible.

Yeah, I’m aware. I was assuming some sort of aligned starting state with atoms whose configurations were perfectly known but I guess that’s not really possible.

If you're thinking "yeah but what if we really knew the starting configuration" then you're not really understanding the uncertainty principle. There is no hidden state that we just don't have access to.

Yes, that’s the realization I came to as well.

You’re one of today’s lucky 10,000. Even Einstein had trouble believing that quantum must be random, but it is. The EPR paradox proves that.

There's a chance - however small - that WE live in that simulation.

More likely that if it's even possible, the chance is near 100%.

However, it actually doesn't matter in any way whatsoever.

Except that we are living in a simulation, then it is morally wrong to be boring

wat

If we are very dull, then we might get switched off

you have no idea what the purpose of the simulation is.

turns out the purpose was for them to study how the borg collective spreads across galaxies, and when we wipe out the borg we've ruined their experiment and they just pull the plug, because none of it was about us. oops.

The purpose of the simulation is always to learn something new.

The simulation theory is indistinguishable from any religious explanation.

Things are weirder than that. It's also indistinguishable from the "brain in a vat" theory, which everyone has heard of and which nobody finds interesting. It escapes me why anyone finds "simulation theory" more interesting than that.

Solipsism isn't even refutable.

It can be from a secularization point of view.

People have rejected religion, and simply invented another one to replace existing ones.

People who believe in the simulation also tend to try and influence it (Scott Adams does), using a method indistinguishable from prayer.

And the Devil Will Drag You Under by Jack L. Chalker https://www.goodreads.com/book/show/584395.And_the_Devil_Wil...

https://en.wikipedia.org/wiki/And_the_Devil_Will_Drag_You_Un...

> In the year 2079, a human attempt to mine a nearby asteroid instead results in the asteroid being directed towards Earth. As catastrophic natural disasters envelop the world, Mac Walters and Jill McCulloch find themselves in Reno, Nevada. Entering a nearby bar, they encounter a demon named Asmodeus Mogart, who tells them that Earth is but one of thousands of universes set up by the mysterious Department of Probabilities, and that he can save the world if he can gather six magic jewels.

Then can I ask the owner of whatever device we're running on, to add one or two zeros to my bank balance?

Hopefully you're not overdrafted when they come through on that request.

The hack is to find emergent behavior so that you don’t need to simulate every atom. That’s what we did with classic thermodynamics. Extremely simple rules that accurately describe crazy complex systems.

From a few years ago - https://www.nytimes.com/interactive/2021/12/01/science/coron...

> To better understand the coronavirus’s journey from one person to another, a team of 50 scientists has for the first time created an atomic simulation of the coronavirus nestled in a tiny airborne drop of water.

> To create the model, the researchers needed one of the world’s biggest supercomputers to assemble 1.3 billion atoms and track all their movements down to less than a millionth of a second. This computational tour de force is offering an unprecedented glimpse at how the virus survives in the open air as it spreads to a new host.

> ...

> Once the virus was loaded into an aerosol, the scientists faced the biggest challenge of the project: bringing the drop to life. Dr. Amaro and her colleagues calculated the forces at work across the entire aerosol, taking into account the collisions between atoms as well as the electric field created by their charges. They determined where each atom would be four millionths of a billionth of a second later.

> To carry out this vast set of calculations, the researchers had to take over the Summit Supercomputer at the Oak Ridge National Laboratory in Tennessee, the second most powerful supercomputer in the world. Because the machine was in high demand, they could run their simulation only a few times. “We only have so many shots to actually see if we can get this thing to actually fly,” Dr. Amaro said.

> The first run was a disaster. Tiny flaws in their model caused the virtual atoms to crash into one another, and the aerosol instantly blew apart. “It basically explodes,” Dr. Amaro said.

> After half a dozen rounds of adjustments, the aerosol became stable. The researchers ran the calculations all over again to see what happened inside the aerosol an instant later. All told, they created millions of frames of a movie that captured the aerosol’s activity for ten billionths of a second.

-----

Aside: the "four millionths of a billionth of a second" is 4fs or 4x10^-15 seconds - https://www.wolframalpha.com/input?i=four+millionths+of+a+bi...

> 0.4 x time for a typical water molecule vibration

The corresponding press release: https://www.olcf.ornl.gov/2021/11/16/we-know-covidisairborne...

The abstract presented - https://sc21.supercomputing.org/proceedings/tech_paper/tech_...

Preprint of paper that matches the abstract mentioned in the ornl press release - https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8609898/

None of the videos are available for me on that page...

Reply on news.ycombinator.com