It's cause of time. Some positions are not very concrete, and you just have to lose your fear and play. In fact, fear is probably what he meant. Fear is such a critical concept in chess its surprising we dont try to understand it more often.
Interesting remarks by grandmasters

@Aquarius
Yes, I supposed it might have something to do with Naka infering how So felt about his position, and how that would affect his play and his defensive resources.
That which you mention about fear is exactly why I love Chapter 4 of Jeremy Silman's: "How to Reassess your Chess".

@Band_Sagger
Sorry I was refering to Aquarius.
Yes I suppose in the end it all comes down to play the position you get. Regards!

@CensoredReality
I think Nakamura proved you right after signing such a beautiful scoresheet just a moment ago!

This is why I recommend changing classical chess time control to game in 60 with a 30 second increment beginning at move 20, to keep the endgame from being a fling fest. The whole idea of the faster time control being to get the top guys to playing chess over the board instead of relying on home preparation, which is what it seems about 80 % of their moves are currently. But everyone tells me I'm nuts, lol.

Wesly So seems like he works himself into a panic. I believe what Nakamura meant is that his own position wasn't good, but if So stopped to try and figure out why, he'd get scared and lose. If So kept playing faster, it meant he knew he had nothing to be afraid of.
'Prep'
'prepared'
'preparation'
all sugared ways of saying "memorization"
Recalling from memory is usually faster than calculating on the spot.
If he plays fast in a complex position, may indicate he is recalling from memory, which indicates he's still in a memorized position, which means you might be in trouble.

Now I'm confused though. GalaxKing seems to imply that prepared moves take longer to thing about and faster time controls would eliminate preparation? Isn't it the opposite... If you already memorized a move, you can play it without thinking about it... Otherwise you need to think and spend time calculating. I dont see the logic.

GalaxKing I think that the amount of time that players spend to recall their prep is not as much as they use while analyzing and calculating lines once they are out of their preparation, and that those long time controls allow the games to have the quality that chess deserves, so I am not very thrilled by your idea, although I wouldn't say you are nuts.
IMO Quiksilverau hit the nail on the head about what Naka meant by the time of So moves, that it has to do with home prep rather than with fear, although fear is a factor as well and it might have had some influece in this case too.
And yes bb_gum234, even Seirawan mentioned in a previous broadcast the irony and sarcasm of many of these top GM's in their commentaries and how they are self-deprecating sometimes without good reason, so that's a good point as well.

Petrosianpupil who is this master you speak of?
It is comforting to read that distrust to some computer evaluations is somewhat ubiquitous amongst strong masters.

It's unfortunate we now live in an era of computers with regards to chess. I was reading a thread posted here about a week or so ago that consisted of the title about how old GMs from the early 1900s or whatever wouldn't be so strong now (something along those lines). But the thought process of these GMs of that era (for example the book "My system") are still used today!
I have more respect for the GMs then than I do now. Give it x amount of years and every game will be a draw at top level with regards to computers. Where is the justification in that?

Naka talked more about his comments in the post game interview. OP and others might find it interesting: https://youtu.be/YbwTBF_ktD4?t=3h45m50s

Petrosianpupil thanks for sharing all that valuable info with us.
amilton542 did they say why they think that GM's from the early 1900's wouldn't be so strong now? I fail to see the logic on that. And yes, as Petrosianpupil says, ever since Capablanca the end of chess due to draws has been predicted and we still seem to be far from it. Psychology and physiology doesn't affect computers, but fortunately they do affect us.
DrinkingLikeTal thanks for that. Naka did clarify a lot to me with those brief sentences in that interview.

James Jackson, He is an IM and only graded 2367 so not strong in comparison, he tells me his game with Giri was short lived. But he is improving. I am trying to follow his advice and not look at computer evaluations or even move any pieces when going over my games but do everything in my head first and try to mentally picture moves more, and only when I am as sure as possible my decision is right then look, but i just cant resist checking my ideas. Anand explained that often computers over score positions when a side dominates lots of squares and space, even though a super GM knows that those squares and spaces are not the critical one, especially when those squares are in the oppositions half.
Jackson is an FM, not IM, 2368, strongest player in Northamptonshire. He comes to the U11 training sessions. I've played him four times!! (and lost badly in three. We didn't finish the other one in an equal position middle game.)
It's unfortunate we now live in an era of computers with regards to chess. I was reading a thread posted here about a week or so ago that consisted of the title about how old GMs from the early 1900s or whatever wouldn't be so strong now
Mark Dvoretsky took players to task in his analytical manual who claimed they would tear Emmanuel Lasker and the older champions apart in his analysis of Edward Lasker vs Emmanuel Lasker in 1924
The type who claim they can beat young Capablanca in his prime the ones who say oh I can beat Steinitz easily and Rubinstien? piece of cake or how about the guy on chess.com who claimed awhile back that a 10 year old could beat up Adolf Anderssen and Paul Morphy? really??? I think they deserve a little more respect than that.
So, I've seen two recent interesting remarks by grandmasters that intrigued me:
First, as reported here on chess.com, Anish Giri discussed with Hikaru Nakamura his round 5 game of the ongoing Sinquefield Cup, and said that because of a confusion with his preparation Naka played 23.Rd3, which is a move that he normally would not have made.
"Sometimes the computer moves are so sophisticated, if you try to understand them, it would take a year.", quote from Giri. Staff member FM Mike Klein then clarifies the quote to us, explaining that some memorized moves are trusted blindly by top-level GM's due to their cryptic nature.
What?
Are top-level GM's playing moves in their openings/early middlegames without understanding them? Isn't that a trait of an amateur approach to openings and to the game in general? How powerful is the influence of engines to the game if even top-level GM's are tossing out their moves relinquishing their understanding of such moves?
And even more recently, in round 6 of the same tournament Nakamura just used the confessional booth to share his toughts on his KID game against So, and said: "basically if So keeps moving quickly I'm going to be in trouble; if he starts thinking I think I'm gonna do very well in this game. So, based on how much time he uses on the next couple of moves we'll see if I'm okay or in a lot of trouble."
What did he mean by that?
Could the speed with which So were to make his moves be an indication of the strenght of his play? I know that amateurs tend to get intimidated or provoked if his opponent makes quick moves, but can grandmasters share this delusion as well up to some extent?