We all know Elon Musk as the funniest man alive.
but did you know that Musk is also a lifelong gamer? for example, here he is in the chess club of his whites only apartheid prep school.
grew out of that though.
really want to highlight the phrase “understandable when all we had to play with were squirrels and rocks.” what the fuck is he talking about?
in any case, he clearly has some issues around the game and I suspect his performance in the chess club. the dude loves to downplay the complexity of the game.
he knows how to tune it up, though.
hmm. fog of war, tech trees, randomized starting positions. that actually sounds a little familiar. could it be…?
OH YEAH BABY! you knew it had to be polytopia!
what the fuck is polytopia?
this is polytopia:
sophisticated Gamers will notice three things here:
-
this is clearly a mobile game first, meaning it was developed to fit into the industry’s nastiest shithole: the ios and android app markets. i’ll tell you right now that polytopia doesn’t have gambling in its microtransactions (which instantly makes it cream of the crop), but you can expect it to be braindead and shallow.
-
this game is for children. look the bright primary colors, the character’s huge eyes, and the low poly style, which evokes minecraft.
-
this game looks a LOT like civilization. specifically, it looks like civilization 3:
For those not familiar, the civilization series is the standard bearer for the 4X genre. it’s a style of slow, macromanagement focused turn based strategy focused on the “four Xes” of exploration, expansion, exploitation, and extermination. thematically, they present societies as machines for producing imperialism. the civilization games are also fearsomely complex and famously deep.
not so with polytopia:
so yeah, it’s baby civilization. but the thing that really kills me is that musk is obsessed with this game. like… really obsessed, according to a story published on yahoo finance.
This puff piece is sourced from “Benzinga,” a tech finance news site that seems to mostly post about crypto, ai and tesla. context clues suggest the content is entirely ripped from walter isaacson’s biography of musk. I have to emphasize that all of this is intended to be flattering - right in the middle, they take a break from the content to shill some affiliate link I won’t click on:
that’s the image of musk they want to convey. anything they include is in support of that.
saying “I am just wired for war” because you play a lot of polytopia is classic elon cringe. beating a game’s developer is nothing to brag about, by the way. most game developers are terrible at their own games.
i would argue that ragequitting to the parking lot to smack down your employee/mother of your children in a baby game is also not something to brag about:
flexing on mom almost shows up more than once in the article…
… except, twist, for some reason he beat his employee but not his partner???
and yes, he sulked like a little shitbag he when he lost. one has to suspect that if his sycophants weren’t constantly throwing polytopia matches against him he’d have turned on it just like he did chess.
before you read this next bit I want to emphasize again that EVERYTHING IN THIS ARTICLE IS INTENDED TO BE FLATTERING.
his brother seems to feel the same way
(and from another benzinga sourced yahoo finance story):
here’s what i think: elon musk isn’t learning shit from polytopia. he’s a profoundly stupid and childish man who likes moving brightly colored blocks around and being told he won. he’s projecting the disgusting and wretched pile of deficiencies which is his personality onto a game for babies. as someone who only fails upward and cannot stand to fail, it is impossible for him to learn from reality; doing so would require facing what is too painful to face.
That’s fair - I guess that’s why I said “it seems” as I can’t confidently state something whose unknowns I don’t fully understand.
But yeah theoretically there might be a way to compress the final output into a physically possible form. Still considering the magnitude of the numbers we’re talking about - this would still be a mind boggling undertaking.
I guess my point was more around elmo confidently stating that chess is solvable and using the actual complexity of the task as a way to highlight how being that confident in such a statement is absurd.
Then again maybe I fell in a similar trap so thanks for pointing it out!
Nowhere near to the degree you seem to think. If you want your computer to output a number in binary that has 10120 digits, it sounds completely ludicrous, but if every digit is ‘0’, then a computer program to generate that number will take up virtually no space at all relative to storing each letter directly.
The algorithm is just outut ‘0’ 10120 times and then halt, so the only difficult part is storing the 10120. (This is trivial for literally 10120, I’ve already described it in six characters counting the exponentiation operator, but let’s imagine it’s actually a similar-sized number that’s less friendly.) The number can be described in log_2(10120) = 120log_2(10) bits, which is somewhere between 360 and 480 bits and can probably still be compressed more. So you’ve represented a binary number with 10120 digits with a program that says repeat the digit ‘0’ and stop after you’ve done it x number of times, where x is a specific number that takes up like 400 bits of storage, probably less. That program can fit on a floppy disk, while the original 10120-bit string itself is a bit too big.
So that’s an extremely simple example, but it shows that there exist 10120-bit numbers that can fit on a floppy disk.
More generally, a massive amount of strings with 10120 letters will have some pattern that can be exploited by a program requiring relatively minimal storage. The ones that don’t have a very high degree of randomness, but that’s extremely unlikely (I’m inclined to guess impossible) to be true of a string solving a real-world problem like chess games. Finding such an algorithm for your string though could be a herculean task.
I edited this comment about 500 times to make it readable and I feel inclined to apologize to everyone who saw this before it was done.
if I’ve followed the comment chain here correctly, the problem remains that you first need to compute all possible game states, which is probably impractical, right?
and you still need to save the data for computing the numbers’ compressed version, or is there a way to compress the numbers on-the-fly that i haven’t heard of?
You don’t necessarily need to. There’s a possibility of defining instructions for winning without that. The idea may just be hard to imagine because chess is complicated.
You can make up stupid examples of games that defy this. Maybe we dream up a game with an ungodly number of states that has an action X that is an available option every turn. If the winning strategy is just “player 1 chooses action X every turn”, you can imagine there may be a way to show that’s true without needing to simulate every single state, and there’s certainly a way to easily store and communicate this strategy.
For a ludicrously dumb win condition, maybe the first playet to press X 10,000,000 times wins, in spite of having lots of other options each turn which introduce lots of other gamestates. You should be able to show that pressing X every turn is a winning strategy without computing every potential state. This can be true of more complicated games too, it would just be much less obvious. Just because no one has conceived of a way to do that yet with a game such as chess doesn’t mean no one ever will.
Hey look, you found another one of Elon’s favourite games.