John Carmack talks

You literally have no idea about programming, right?

he speaks like an autist im sure every othet tech professional isn't that casually into higher math

I develop game engines, my job is 80% mathematics, 15% pseudocode, headers, and actual C++, and 5% listening to "coder" drivel about type qualifiers, linux portability and other trash. You can only make something productive and performance-intensive with algorithms if you understand the math behind them.

You could have just said "yes", most programming has no relation to mathematics. It's something like carpentry and to be a carpenter you do not need to know the material science of wood.

>my job is 80% mathematics
Just repeating basic geometry again and again?

A carpenter cannot make a chair that takes 150KG, doesn't wobble, uses the minimal amount of mateiral, and lasts 20 years, without mathematically proving his design is such. I can glue together several PVC pipes and sew old jeans on top but that doesn't make a good chair.
>no relation to mathematics
Programming itself is a subset of mathematics, it's your problem if you can't describe your actions in the world's most verifiable mutual language. I laugh at programmers who make changes and benchmark and write shitty unit tests, to prove their design choices were correct. I can write a very fast bubblesort myself but that's useless.

>chinese cartoon watcher
>closeted pedophile
>dignity

>If you have some dignity
And that's where I get you, friend. Heh

>A carpenter cannot make a chair that takes 150KG, doesn't wobble, uses the minimal amount of mateiral, and lasts 20 years
That is false.

>without mathematically proving his design is such
What does that even mean? You can not "prove" a design. An engineer calculates estimates.

>Programming itself is a subset of mathematics
It isn't.
It is telling a computer what to do, most which is really basic repetitive stuff.

>it's your problem if you can't describe your actions in the world's most verifiable mutual language
???
Most mathematical concepts can not be expressed computationally, as evidenced by the most fundamental object in analysis, the real numbers, being almost always not computable.

You seem extremely confused about what mathematics or programming actually is.

>That is false.
>calculates estimates
Yeah that's why "ye olde stooles" weigh 50kg and are thick as fuck, yet modern IKEA can be slammed into the ground no problem.
>It isn't.
Literally everything you do is morphisms over a set of uniquely addressable objects. Programming has existed before arbitrary combinations of transistors, before random access memory, before computers themselves became a physical object. It was all done and proven correct via abstract algebra, type theory, category theory, etc.
>basic repetitive stuff
Maybe when your entire job position consists of "you are given a header file and pseudocode. Write the source code and some unit tests.". My job position is to supply these in the first place.
>can not be expressed computationally
I can apply a predicate over all real numbers or implicitly define every pair of real numbers with sum 5 (or any number). Want to see me do it?
>actually is
What are your credentials actually? What do you do and define yourself as?

DMT? DMT.

You are just making stuff up, right?
You have never programmed anything in your life...

>I can apply a predicate over all real numbers or implicitly define every pair of real numbers with sum 5 (or any number).
Missing the point so hard.

>Literally everything you do is morphisms over a set of uniquely addressable objects.
You are retarded. Thoroughly retarded, just because something is describable through mathematics doesn't make it mathematical.

>What are your credentials actually?
I have a degree in mathematics and actually programmed something not trivial once.

>Missing the point
You said something and I showed you're wrong.
>doesn't make it mathematical
I clearly showed you how programming was another name for applied mathematics since before children clicked on a keyboard and pressed F5. The only reason you can even run anything on your computer that is not native machine code, is because a compiler (correctly) created such a homomorphism.
>I have a
Cool, what discipline? What did you program?

You are so wrong it’s amazing. Programming is a mathematical discipline which involves both the theory of creating/choosing the correct algorithm, verifying its correctness, and then finally coding it into the appropriate medium. With your “math degree” you should realize that part of programming is the required computer science. This means mathematics to both represent and prove new concepts beyond any doubt.
I have a feeling you are hanging on to some weird presumption of the argument in order to technically excuse yourself. Either way, the other user has thoroughly fucked you.

>I clearly showed you how programming was another name for applied mathematics
By that standard LITERALLY everything is mathematics and "mathematics" uses all meaning.

>The only reason you can even run anything on your computer that is not native machine code, is because a compiler (correctly) created such a homomorphism.
But it isn't a homomorphism, the most important reason is that the mapping is not into itself and doesn't preserve operations.

>Cool, what discipline?
Usually analysis. Wrote my Bachelor's thesis about something in numerical analysis.

>What did you program?
From small games to implementing lots of algorithms, some physics simulations, 3 Body problem, etc.

the clashing of 2 of our generations greatest minds
fuck yeah ill watch that shit
upvote too

>mathematical discipline which involves both the theory of creating/choosing the correct algorithm, verifying its correctness
That is computer science, nothing to do with programming.

>and then finally coding it into the appropriate medium.
That's programming.

>With your “math degree” you should realize that part of programming is the required computer science. This means mathematics to both represent and prove new concepts beyond any doubt.
I am not denying the existence of CS, which you seem to describe here.
But programming is the act of giving instructions to a computer. A programmer might also be a computer scientist, but most programmers write programs which have almost no relationship to mathematics.
If you don't believe me, enter /wdg/.

>Either way, the other user has thoroughly fucked you.
He has made only non-sensical arguments this far.

>LITERALLY everything is mathematics
Yes, that's the final red pill. The pink pill is "Everything can be explained via mathematics".
>not into itself
You're thinking of automorphism?
>preserve operations
The C standard clearly defines what integers are and what happens when you apply operations on them. As long as you don't cause undefined behaviour, the compiler guarantees adding two unsigned 32-bit integers in 2's complement around in registers on a supported platform, is the same as adding two u_int32_t's. Why? Because group homomorphisms. {0, 1, 2, ...} mod 2^32 is a group with addition, and C and registers provide such compatible numbers and operations to work with.
>numerical analysis
Now we're talking. What topic?
>implementing
"Given pseudocode, ..."
Okay

>Yes, that's the final red pill.
If a single concept encompasses everything, then it is meaningless.
Me picking my nose is mathematics, as much as doing my PDE exercises.

>You're thinking of automorphism
Yes, but it isn't structure preserving either.

>The C standard clearly defines what integers are and what happens when you apply operations on them
NOPE, for example the C standard doesn't define how man bits exactly are used by an int.
It is NOWHERE defined what exactly happens in many cases, overflow us even undefined behaviour.

Anyway, addition in code is not equivalent to addition in executed code, if I replace two numbers in source code by their sum the result can be different from storing these numbers in ints and adding them.

It doesn't preserve group structure and the mapping isn't even well defined by the standard, sorry.

>"Given pseudocode, ..."
Usually I was given a purely mathematical description, but yes.

>What topic?
This goes obviously in an area where I can doxx my self, but it was about the properties of a certain infinite dimensional operator which describes a certain physical phenomenon and you could numerically bound it's norm.

Your brain does an incredible amount of calculations to put your finger into your nose without looking at all, just from memory. No self driving car has been demonstrated to go over a known path and still evade obstacles without any guidance at all, LIDAR or anything.
PDEs are very slow for you to do because you explicitly write things down and think about formulas you've learned, mostly by photographic memory. However, your brain automatically approximates indefinitely large systems of PDEs to stimulate all the muscles to just the right tension, to lift up and bend your arm, fitting that fingertip inside the nostril on the first try. And it works just as well in strong wind, in sunlight, when one nostril is clogged with goop, and in space. That is almost magical to me.
>isn't structure preserving either
I don't understand this. The entire point of a compiler is to correctly make the computer manipulate variables in RAM the same way your functions manipulate variables on the address space. Granted, no compiler has been formally verified, they do an excellent approximation at least.
>how man bits
Bits are irrelevant, you are given a word, and work on words. How long a word is is irrelevant.
>what exactly happens
You can literally see the standard, there is no such thing as unsigned integer over/underflow in ANSI C or ISO C++. It's totally fine because they're treated as remainders.
>can be different
The standard says it can't in this case for integrated operations. Otherwise, there would be a supported architecture where pre-computed sine tables and runtime sine computation would yield different results, just as an example. But no such has been reported.

so what time they talk about psychedelics?

my guess is carmack asked in advanced to avoid this shit

>Dude, I played Quake on a T1!
>I played Doom all day
>Yep, Quake was a good game
Joe is cringy as fuck. It's obvious that he has no clue about what Carmack is talking about. I was hoping for a multi-hour session of pure
reciprocating technobabble.

[B] : √
[R] : √

>I develop game engines, my job is 80% mathematics
You know how I know you don't develop game engines?

Based cumbrain.

All of these are fine except for
> small integer
> macroscopic
Jesus cringe. Talking way too fast, this is what occurs. I'm not watching 3 hours of this. I have watched him in the past and have gained a lot more understanding about people in the industry. TBQH, and I do work in silicon valley, this 'rockstar' programmer hype is just that. Also, most code is not complicated. The constructs and paradigms are just convoluted and constructed by autistic making it a pain in the ass to deal with.
> Graduate top Uni in actual comp eng.
Sorry, he sounds more like a drop out who overcompensates because he wasn't formally taught.. Oh look I am correct.
Correct
False. The biggest sham of tech is this. You rarely use high level advanced math on a day to day basis even for algorithms... why? because some crackhead PhD already did so in Uni and proofed all of the Algos you use at a corp. It's just copy/pasta in some corporate library. Many of them even have license statements because they are pasta.
> mfw brainlets who don't work in tech think there's magical rockstar coders
LMFAO
Bullshit. 30% of your job is math (basic geometry and matrix math) - Hello highschool. Just about none of it is calculus or advanced math. 20% is pseudo code, 30% is autistic 'domain' specific garbage because incels never humanized game engine dev and the other 30% is C++/OS horse shit.

>incels never humanized game engine dev
what does this mean

Game engine dev is hard because autistic incels never formalized or made sensible the grinding aspects of it.

Carpentry is a shitty analogy and far more niche. Most things programming related are a grind because incels and autists are unable to translate domain specific paradigms in a straight forward way, suck dick at documentation, and have a perverse desire to make things overly terse/complex so they feel elite. Prove me wrong, you can't.

Unit tests.. most engineers suck dick at.
Math has fuck all to do w/ anything in modern systems because everyone just tosses more hardware at compute. At best your doing basic +/-/x/division which any retard can do. Even in the embedded world. Again, you prove my point about autistic faggots trying to make everything seem so complex.
> I can write a very fast bubblesort myself but that's useless.
do you know how long it took to develop this algo, proof the run-time in the original paper? Which is why its just copy/pasta'd.

Oh and Big oh' notation is largely retarded because it gives zero considerations to mem access which is what usually kills you.

>Game engine dev is hard because autistic incels never formalized or made sensible the grinding aspects of it.
You have no idea what you're fucking talking about

>You are so wrong it’s amazing. Programming is a mathematical discipline which involves both the theory of creating/choosing the correct algorithm, verifying its correctness, and then finally coding it into the appropriate medium.
^everyone look at the faggot whose never worked at a tech company with a billion dollar+ market cap. No one does this. If it runs it ships.
> tfw I've literally witnessed software with over $8 billion in sales fail in customer use case tests before shipment and everyone scramble because the shit wouldn't even boot
Guess what it was, shitty memory management.
What math exists for that?
> algorithm performance
LMFAO, if it runs it ships.

You don't know shit. There's maybe 10 people on a 1000 person team who copy/pasta algos into a corp library who have to do basic mathematical performance analysis. The rest just call their APIs. those algos are standard bullshit they teach freshmen year in college. Take a look at linux kernel algos .. it's basic shit

actually, what i find tolerable about rogan is that when he's intellectually outclassed he mostly just shuts the fuck up and lets his interviewee almost dictate the conversation.
there are so many brainlet interviewers that don't shut the fuck up

John Carmack is the same to technology as Seinfeld to comedy

Programming is at best some combination of formal logic and language. Calling it math is absurd.

>formal logic
-ENOTTURINGCOMPLETE

I thought Carmack killed himself after Jobs died.

A technically proficient master?

I gotta say, I hate podcasts, but listening to joe rogan talking to alex jones is fucking amazing.

who's the bigger idiot?

This is why management gets paid more than you fools. Understanding math is easy.

Wrangling and making autists go back to their desk and do all the work for you, now that's real science.

Either this, or they just have a small talk before the show to select topics.
Because if Carmack doesn't know shit about DMT or whatever else, there's not point in asking during the interview.

you

Nice blog post.

I mean come on whats the big idea.

Attached: 1564501277875.png (495x498, 327K)

did joe ask him about stealing computers as a kid? a brain with legs?

You're speaking to one... Someone with tons of coding experience. You probably need 10 autists max for a 1000 person engineering team who are algo/math nuts (if need be). You have them code platform agnostic libraries for algo shit and you have your next line of engineers leverage it through apis. Typically though, it's all basis linked lists/hashes/graph bullshit that doesn't require anyone serious in math/algos.

Oh i do. OpenGL is one of the most convoluted languages known to man. Vulkan finally adds some normalcy but even then. Game engine's aren't that serious either. Threads/processes/locks/syncs/memory management .. Your basic undergrad who passed OS can work their way into it. As for the industry et large, like any painful engineering discipline, it's usually do to no one stopping long enough to formalize aspects of their work. Now-a-day people just leverage unity/etc and ship it if it works.

>Oh i do
>OpenGL is one of the most convoluted languages known to man.
Uhhhh

Attached: 1501576731663.gif (300x290, 1.74M)

Correct.
It means autistic faggots who enjoy convoluted/terse paradigms never met with language experts/sensible people that formalized the 'grind' details of their work into more readable/understandable constructs. Thus, the hardest part of doing their work is wrapping your head around the shit tier conceptual framing, languages (like OpenGL), etc. If developing anything requires fuck tons of math beyond 'science' related work, you're fucking doing it wrong. A physics engine one can understand but guess what, they were formalized and now its just copy/pasta and API calls.

Do you have down syndrome? lmfao

COPE