Serious question: If a highly intelligent being, better than humans, is about to code something, how would they probably do it?

Will they use the same concepts like control flow, iterations, types, operators, object inheritance, etc?

If they are quantum capable, how can they code with booleans when it can be both true and false at the same time? Will they code truthy and falsy with another dimension like time-space temporality?

Do their code simultaneously modify the hardware or bio-hardware as it iterates over the outcome of the code?

Does input and output even relevant to them?

How do they represent infinites?

Do they have similar github workflows or they telepathically update the source code?

Do they embed their program in their DNA? Then pass to offspring the codes they already created?

Do they code using a language or do they use some frequencies and material science that simultaneously show real world output?

And do they have their version of devRant?

We probably can't comprehend it

Dude, are you high?

I mean, quantum bits are less a spectrum, more a composition of possible states. Quantum computers aren't that far fetched.

Before a pedant comes in, we already have quantum computers but they're very restricted. I also have a physics degree.

On the whole quantum bits thing, read into number factorization. It's less 1/0 as we currently use them, more like voltage and resistance. Bits settle into acceptable, preferable states with lower resistance. The resistance is caused by the problem definition.

Separation of concerns is almost certainly a universal principle, even evolution does it to some limited extent in environments where flexibility is valuable (which is most environments we can visit).

Boolean algebra is really useful specifically because its simplicity lends itself to mathematical reasoning. Although they would probably ponder questions far beyond our understanding, the utility of this tool is unlikely to be limited so they would probably have something similar.

Type theory derives directly from set theory which is the foundation of mathematics. It's less than a hundred years old so claiming that a being far more intelligent than us would also build their mathematics on set theory is a bit brave, but it's certainly a very good foundation.

The complexity of clean code grows geometrically, because the number of relations is kept artificially low. The complexity of messy code grows exponentially. A being far more intelligent than us would probably have the same or similar architectural concerns as we do, just on a far greater scale.

Uhmm... Yes?

Dunno what drugs you took, but I'll take the hypothesis bait.

Look at star trek in it's entirety and science fiction in general.

You will find quite a lot of simulations which might or might not be to your liking.

In a nutshell, it all boils down to form of intelligence and it's comprehension of the universe.

E.g. the Dominions as an entirely anti individual species vs the secrecy and xenophobia of the Romulans.

Star Wars Ewoks are an entirely different thing. I guess the teddy bears didn't really care for IT.

All in all, and I give Babylon 5 the credit here where the hundreds of different religions of the human earth were honored, we humans always have one thing in common: diversity.

Some people see science as a well defined building with a concrete foundation, while it's in my opinion rather a waka that sails on very stormy water.

A lot of our understanding stems less from facts - more hypothesis accepted as valid till proven otherwise.

And our diversity is the thing that keeps the flow of assumption, validation and falsification going.

Look at e.g. your example of using DNA as a storage. The process of DNA discovery and unfiddling it's mysteries took 150 years and is still ongoing.

The shift in human mindset during this time was a pretty large one.

Wait another 100 years, if we have 't bombed or otherwise erased ourselves from earth, it will be interesting to see what we might have then.

Idiocracy is currently most likely, but maybe it goes all another way.

Sounds like machine elves. Look up starpilot33 or 434 on YouTube.

True and false would exist but also undefined (indeterminate) and closer to true and closer to false. It would be more like a scale or spectrum.

Even as complexity increases the basis upon which it is founded remains the same.

Lol devrant for space aliens

"Serious question: If a highly intelligent being, better than humans, is about to code something, how would they probably do it?"

there's so many things that can not be told about this scenario. but there's one i'm absolutely sure of:

they won't be using javascript.

I work with some. They don’t do TDD and they test after they deploy to prod.

@TrevorTheRat "deploy to prod" - that's so basic. my first employer was "code on prod" - in combination with "what's version control?"

I know what your goal is and It's a really good question, but if we knew the answer we would already be doing it in the first place.

I tried thinking of new programming paradigms a couple of times, but the reality is that such thoughts come only about once a century. Not to mention a lot of our programming paradigms are limited by our hardware design. It needs to execute sequentially anyway using cpu supported control flow instructions. Chances are that if you want to invent something revolutionary, you might need to start at hardware level. Otherwise It's just variations on what we have already.

@Hazarth exactly why I mention about simultaneous hardware-software interactions where the hardware becomes its software at the same time. We do have a static hardwired rope programmed thing that was used in the Apollo spacecrafts. But these are static. Imagine if a biohardware thing kepps modifying its own hardware to be its own software.

The nearest to this idea is our human brain and dna.

If you had a "perfect" AI tester (that knew which output humans found optimal) then you could just have an AI generate code with a genetic algorithm, trying all sorts of whacky possibility until it found the version that scored the highest rating with the AI tester.

Much of what we consider good code is just based on making it human readable and preventing human sloppy mistakes. An AI could rewrite the entire codebase so fast that none of that would be necessary.

However that's Miles away - most genetic algorithms are hand coded to only tweak a given set of variables within a human program. They don't rewrite the parent program itself.

And I'm not sure if it would be possible to have a perfect AI tester - it works for a small program like image compression. But for a large program even if the AI thought "on average this output is best for this website audience" some human would go "no, change the logo" on a whim.