Chess and Code Part 2

This is part 2 of a conversation between Johnny Coppola (johnnycopes.com) and I comparing chess and programming. Enjoy!

Transcript

The following transcript was automatically generated by an algorithm.

00:00 : this example is used to illustrate the
00:03 : point that
00:04 : sometimes you get yourself into trouble
00:08 : and then
00:09 : you dig a deeper hole for yourself
00:13 : yeah yeah and like compiling
00:18 : one way of explaining that is the sunk
00:21 : cost fallacy
00:22 : it's like because i've already
00:25 : committed to this plan
00:27 : even when things are going downhill i'm
00:30 : not gonna get away from it i'm gonna try
00:32 : harder at executing this plan even when
00:36 : the right thing to do would have been or
00:38 : the more beneficial thing to do would
00:40 : have been to cut your losses early but
00:43 : yeah but you still don't do it and um so
00:46 : this is an example in chess where it
00:48 : happened to my opponent and i can gloat
00:50 : about it a little bit um
00:54 : i'm going for this pawn here uh like
00:57 : like he
00:58 : originally he had this pawn guarded with
01:00 : the knife and because he moved the
01:02 : knight away
01:04 : i decided to attack this pawn because i
01:06 : i i thought i thought like the reason he
01:09 : wanted to move this knight here is to
01:12 : trade off the knight over here right so
01:14 : he's going to take
01:15 : and then if he takes i'm going to take
01:17 : with my pawn
01:19 : and i calculated this up to this point
01:22 : and i realized that well if i take with
01:24 : my pawn my pawn is going to be attacking
01:26 : his rook
01:27 : so he's going to have to spend time
01:29 : moving his his rook and while he moves
01:31 : his rook i i will have time to take this
01:34 : pawn so i get to win a pawn at the end
01:37 : sure
01:38 : so that's exactly what happened
01:40 : so this here now i'm like yeah i get to
01:43 : take this pawn now i won my pawn
01:45 : because his rook has to move away
01:48 : but at this point i think he was in that
01:51 : mindset of no i don't want to give up
01:54 : this pawn so oh boy so what he did was
01:58 : move his rope here he want to attack
02:00 : this
02:01 : pawn of mine he's like okay we're just
02:03 : trading ponds because i'm gonna get you
02:06 : take my pawn
02:08 : i'm i'm gonna be able to take your pawn
02:11 : but uh but i at this point i realized
02:14 : no i can attack your rook while
02:16 : defending my phone so you still have to
02:18 : move your rook away
02:20 : so now the next move you're gonna move
02:22 : your rook away and then i'm gonna get my
02:24 : pawn
02:26 : and and that but that's not what
02:28 : happened he said no i'm gonna guard my
02:30 : rook with my rook what it allowed me to
02:33 : do is uh block off the connection of the
02:36 : rooks and i'm winning this rook
02:39 : yeah he just dug himself a deeper and
02:42 : deeper hole initially it was a small
02:45 : loss he was going to lost this pawn but
02:47 : it became a bigger and bigger loss as as
02:51 : he uh kept refusing to concede
02:55 : and uh and then programming
02:57 : you could double down on a bad pattern
02:59 : or
03:00 : anything like a concrete example i've
03:02 : seen in this case he was trying to hold
03:04 : on to that rook
03:05 : no matter what sorry hold on to the pawn
03:09 : um
03:10 : and maybe it could be you you try to you
03:12 : try and approach to solve a problem and
03:14 : it doesn't really work but you're like
03:16 : no i'm going to like force this solution
03:18 : through
03:19 : yeah that reminds me so oftentimes you
03:22 : extract
03:24 : some
03:24 : reusable function that you can use for
03:27 : certain things
03:29 : and then when there's problems that it
03:31 : doesn't immediately solve you add
03:34 : different modes like you add an input
03:36 : that is like a mode that goes into that
03:38 : function and by the choice of mode you
03:41 : add some switch statement or if
03:44 : statement inside the component
03:46 : that makes it do slightly different
03:48 : things and then you keep adding more and
03:51 : more modes to that thing and then like
03:54 : the right thing to do is just to call it
03:56 : quits and say this this isn't holding up
03:59 : anymore i should try to find a different
04:01 : approach to it yeah simply adding more
04:04 : modes but i've seen those kinds of code
04:06 : get crazy i have one more that i wanted
04:08 : to tell you as well the parallels of
04:11 : learning both chess and programming
04:14 : languages
04:15 : have like you know not the easiest but
04:17 : it's a reasonable learning curve to get
04:19 : up and running so
04:20 : parallels in programming languages you
04:23 : have learning the basic data structures
04:25 : like variables objects arrays in chess
04:28 : you have learning the different pieces
04:31 : and how they move right you have things
04:33 : like in programming control flow
04:36 : in program in chess you have the game
04:38 : flow you know and the objective of
04:42 : checkmate and chess versus in
04:44 : programming it's typically to accomplish
04:45 : some kind of task
04:47 : however
04:48 : both of them
04:49 : while relatively easy to get into have a
04:52 : tremendous
04:53 : depth in order to master and take many
04:56 : many hours because there are so many
04:58 : concepts and principles and uh lurking
05:01 : beneath the surface
05:02 : yeah chess is very famous for being
05:06 : extremely complex because of the sheer
05:08 : amount of combinations that you can get
05:11 : it's very deep because there aren't all
05:13 : that many things you can do and
05:15 : similarly in programming languages to
05:17 : some extent there aren't all that many
05:19 : structures or pieces you can use
05:21 : but the depth of it is insane because
05:24 : the like the variation of what you can
05:26 : do is just so it says it's nice
05:29 : for study
05:30 : chess is actually more tractable than
05:33 : programming because programming has the
05:35 : subjective element to it yeah both from
05:38 : the product development side and the
05:40 : sort of developer side of things even
05:44 : though chess has a psychological side as
05:46 : well you can reduce it down to the just
05:48 : the choices you can make on the board
05:50 : one thing i've just observed like
05:52 : showing people the basics of programming
05:54 : or like the basics of chess is there's
05:56 : typically this moment where even after
05:58 : they learn like okay this is a function
06:00 : this is an array and like this upon this
06:02 : is a bishop
06:03 : i know what the things are but there's
06:06 : this moment where they're just like i
06:07 : don't know what to do
06:08 : like in chess they're like i i i don't
06:10 : know what to do like there are so many
06:12 : options i don't know where to go or like
06:14 : what i'm trying to do in programming it
06:16 : can be like the same type of thing yeah
06:18 : you present them with a simple task and
06:20 : they're like i don't even know how to
06:21 : begin have you come up with a way to
06:23 : talk to them i don't know what to say
06:26 : it's hard it's hard it's really hard and
06:27 : i think sometimes you have to lead by
06:29 : example
06:30 : um where i think about when i like ever
06:32 : approached my first algorithm problem of
06:34 : like a reverse a string i was completely
06:36 : clueless and i had to like kind of dig
06:38 : into some examples
06:40 : um but i think
06:42 : short of that
06:44 : breaking it
06:45 : helping people think about breaking it
06:47 : into smaller steps
06:49 : really helps you're like well that what
06:51 : about we know that controlling the
06:53 : center is good
06:54 : and if we move a pawn
06:56 : we can
06:57 : create more options for movement
07:00 : and also make strides towards
07:02 : controlling the center in programming
07:04 : it seems like we're going to need
07:07 : a place to put our stuff so maybe we
07:09 : could create an array you know to start
07:12 : things like that like really small
07:14 : trying to train people to think in
07:16 : really small pieces and not get
07:17 : overwhelmed at like the prospect of get
07:19 : the king or like write the whole
07:22 : application
07:23 : so we we have these chess ais that teach
07:26 : humans how to play better like i have a
07:29 : computer acting as my teacher which is
07:32 : really incredible that might be coming
07:34 : for
07:35 : programming maybe i could see that what
07:38 : i want to say is that there's a benefit
07:41 : in calculating to multiple levels of
07:44 : depth when you're evaluating your
07:47 : choices here the computer is actually
07:50 : telling me
07:52 : night here is the best move
07:55 : i couldn't really understand
07:58 : why
07:59 : the only thing i could think of is i
08:00 : know like knights are really strong on
08:02 : the sixth rank okay but otherwise
08:04 : there's no immediate benefit it's just a
08:07 : really good position for the knight are
08:08 : you trying to trade off this bishop i
08:10 : did consider training on this bishop no
08:14 : your your knight is way in a way better
08:16 : position than that that bishop is really
08:18 : stuck in a lot of ways because of those
08:20 : pawns
08:21 : i i thought my move was really good i
08:23 : thought my move was your move
08:25 : up on up but the computer said no that's
08:28 : a blunder that's a horrible move really
08:31 : yeah but i still ended up coming on top
08:34 : my plan even worked out it's like
08:37 : because my plan was like
08:39 : hey i want to win the exchange and
08:42 : except for my plan initially was like
08:44 : i'm gonna fork the rook was here and if
08:46 : i move my knight here
08:48 : then i fork these two rooks and i win i
08:51 : win one of the rooks that was my plan
08:53 : initially and he was like oh i see that
08:56 : so i'm gonna move away the rook i'm
08:58 : gonna move the rook here
08:59 : which they also consider a blender yeah
09:01 : they also the computer also thinks the
09:04 : concert the computer is just better than
09:06 : anybody so
09:08 : what can you say it even makes magnus
09:10 : carlsen humble but but my plan was like
09:13 : hey i'm gonna go here and then what he
09:16 : would have to do is move the rook
09:19 : and now i have to at
09:21 : i can move my knight here i'm going to
09:23 : win this rope or he's going to have to
09:25 : take my knight with that rope i win one
09:27 : of the rooks right and and it worked and
09:30 : it worked so
09:31 : so he ended up taking my pawn and now i
09:34 : get to take my rope and i was i was
09:36 : happy it's like yay my plan worked i won
09:39 : the exchange i won one of the rooks why
09:43 : does the computer think this is the best
09:45 : move
09:46 : you sort of have to
09:49 : see several steps deeper
09:52 : into the position in order to understand
09:54 : it
09:55 : and the reason is if you make a random
09:58 : move for example you take the bishop
10:01 : take the knight
10:02 : this is checkmate yeah
10:05 : it's like holy crap
10:07 : well okay let's say the rook
10:10 : goes here okay and then now you take the
10:12 : bishop and you still open up this file
10:14 : and then you check the king the king has
10:16 : to go here and then what you do is you
10:19 : check the king again the king goes here
10:22 : and then you double up your rooks
10:25 : on the last rank
10:26 : and then this knight is trapped
10:29 : yeah
10:30 : it was like holy crap i like how many
10:33 : moves do i have to see ahead in order to
10:36 : see that that night move back there
10:39 : was the best move
10:41 : i will say though like and i did not see
10:44 : that many moves ahead this is where i
10:46 : feel that like in both programming and
10:48 : chess principles can come in handy
10:50 : because what i talked about before is
10:52 : like i don't necessarily know what the
10:54 : best thing or why it's the best but
10:56 : putting a knight on the sixth rank i
10:57 : know is a smart move
10:59 : and then here
11:00 : what they eventually did
11:02 : doubling up your rooks is always very
11:05 : often a very smart move because they're
11:07 : so powerful when they're together
11:09 : and then so you're like kind of creating
11:11 : opportunities even if maybe you didn't
11:14 : necessarily see it like an obvious
11:16 : opening right away it's like if you
11:17 : bring yourself a step closer
11:20 : then maybe like you're like oh damn i
11:21 : can do this you know it's like these
11:24 : options present themselves and similar
11:25 : to programming if you just like
11:26 : structure things well and set yourself
11:28 : up well it becomes like much easier to
11:31 : execute
11:32 : future things that maybe you didn't
11:33 : anticipate at an earlier stage as well
11:36 : yeah i think that's true it's something
11:38 : that maybe i
11:40 : you know
11:41 : haven't really fully appreciate it but
11:43 : yeah i think that's very true
11:46 : the amateur's mind by jeremy sillman
11:49 : i i got that book and i had like a
11:51 : period where i was studying it and going
11:53 : through the examples and he has like
11:55 : eight or so very broad principles with
11:57 : like tons of examples
11:59 : and uh i just
12:02 : he kind of like point he's kind of like
12:03 : emphasizing what i'm saying as well it's
12:05 : not necessarily about seeing like a lot
12:07 : of moves ahead it's like
12:08 : just knowing kind of like the best
12:10 : practices if you will
12:12 : is it a chess book yeah it's a chess
12:14 : book got it but anyway to you to your
12:17 : original point toby about the deep study
12:20 : absolutely like i as someone i like
12:23 : learned how to play chess really young
12:25 : and like just played for fun like like
12:28 : i've played like a lot of times right
12:30 : and it wasn't until like that that that
12:32 : stint like in the winter of this past
12:34 : year like you know and you were involved
12:36 : with that too
12:37 : i was like i actually like for the first
12:39 : time like seriously studied the game
12:41 : and like used the you know the chess
12:43 : engine to walk through games and stuff
12:46 : and it like it really elevated my game
12:48 : like in a way that just playing over and
12:50 : over again but never like going deeper
12:52 : dead you know just like it advanced
12:55 : things in such a different way
12:57 : um so i i completely agree and the same
12:59 : you can apply to programming languages
13:01 : you can just kind of like get through
13:02 : your day job and like
13:04 : you know maybe like do a little bit here
13:06 : and there but it's i think it really
13:07 : pays off to like spend some extra time
13:09 : to like
13:10 : go deep into like
13:13 : uh like programming language or
13:14 : programming study at a more concrete
13:18 : level if you look at the calculation of
13:21 : chess outcomes as decision trees if i do
13:24 : this they do this then i do this and
13:27 : then but at each fork there's actually
13:29 : multiple fan outs what you're trying to
13:32 : go for is a global maximum like you're
13:35 : trying to go for the optimal move
13:38 : with all of those possible outcomes in
13:40 : mind and the best move is a move that
13:44 : considers all of those possibilities
13:47 : versus like considering the choices at a
13:49 : superficial level so like i feel like
13:52 : one one of the things we might do is
13:54 : like choose this framework versus choose
13:57 : that framework that's like a choice at a
14:00 : more superficial level so so like we get
14:02 : these
14:03 : flame war debates about my fr my
14:06 : framework is better than yours or my
14:08 : programming language is better than
14:10 : yours sure but but
14:12 : that's only the first move of the
14:14 : variation regardless of which choice you
14:17 : make on the first move
14:20 : there's so much more richness
14:22 : that's underneath that
14:24 : for each program that you write
14:26 : there's lots and lots of choices you
14:29 : make every single line of code you write
14:32 : how do you make this choice versus that
14:34 : and if you encounter this
14:36 : what choice are you gonna make
14:38 : we go about those things we maybe follow
14:40 : some some good rule of thumb that we
14:44 : learn but also if you want to make a
14:47 : really good choice
14:48 : sometimes there's just no avoiding the
14:51 : having to traverse pretty deep in all
14:54 : those options first yeah before you
14:56 : realize what the best choice is at the
14:59 : top level you know
15:01 : and that's what you recommend when
15:02 : you're like if you're not sure what the
15:03 : solution is it's worth trying out
15:05 : several and then picking the best one
15:07 : yeah that's what i advocate like if you
15:09 : don't know what's the best code
15:11 : structure it's worth it to try multiple
15:15 : ways to refactor the code i have one
15:18 : lightweight one as a closing thought
15:19 : this is just more of an observation but
15:21 : i really
15:22 : in both like the chess and the
15:25 : programming community
15:27 : i really like how it attracts people
15:29 : from all over the world like you play
15:31 : chess online it's like you never know
15:32 : which country the person is going to be
15:34 : from you go to like a programming
15:36 : conference and you'll have like
15:38 : attendees from all different
15:39 : combinations i just think that's really
15:41 : cool and it's like cool to have that
15:43 : thing in common with
15:45 : uh people that otherwise you know you
15:46 : might be pretty culturally distinct or
15:48 : even like speak a different language
15:50 : yeah i just i really like that aspect
15:52 : yeah definitely wait i like it when um
15:56 : when i'm playing chess with somebody and
15:58 : they i see their country flag
16:00 : there me too at the conference usually
16:03 : it's not as international depends on the
16:06 : conference you got maybe i should go to
16:08 : europe like the euro js con it's true
16:10 : because like in chess.com it's like you
16:12 : don't even need to speak english at all
16:14 : like it's and i'm sure many people don't
16:16 : but the conference you know typically
16:18 : like the language is going to be
16:20 : the common language would be english
16:22 : typically attending to be from like
16:23 : wealthier countries trust is really so
16:26 : universal that you can you can get to
16:29 : know somebody without being able to
16:32 : speak to them yeah real life i traveled
16:35 : to bosnia
16:36 : and i was like waiting for a family
16:38 : member to show up for an afternoon i was
16:40 : like walking around by myself
16:42 : and there were a bunch of old bosnian
16:44 : dudes
16:45 : in the park
16:46 : playing like a large version of chess
16:48 : and i was like able to stand by and like
16:50 : watch it
16:51 : and i was like i don't know anything
16:53 : they're saying
16:54 : you know they're such like a different
16:56 : age bracket than me like all that
16:58 : we're very different but i know exactly
17:00 : what's going on in this game