If I go solo, I'm doin' a song with Bolo
A big Chinese ***** screamin', "Kuniva, yo, yo!"
Do you think humans will ever be as good as robots?

I presume you meant will people ever be as strong as computers at Chess. Not unless we have computer implants.
But in the recent Alpha GO tournament that defeated Lee Sodol professionals expressed awe, amazement, excitement, sadness,and a love for the game. The AI showed none of this. Nor did it say Iḿ bored could I do something else?
The other thing to remember is games are designed to challenge and entertain, people not machines. They challenge our limitations. But we are conscious and have spirit.
If I go solo, I'm doin' a song with Bolo
A big Chinese ***** screamin', "Kuniva, yo, yo!"
Posting under the influence can be hazardous to your health 😂

i guess humans are the ones who are forcing robots to study chess 24/7 and calculate a bunch of stuff for humans, but humans are not studying what they've been making the robots discover

No, engines are improving much faster than humans and there doesn't seem to be much of a slowdown yet.

Robotics can be integrated into human organic body. It's called Cyborgs or Bionic men and women. Metabolism would be like Normal human. Reproduction similarly. A human can qualify for bionic enhancement starting age of 18 or the legal age of his abode.
In fact human can also be genetically engineered as super human. ( The same as genetically modified foods that are 2x to 10x bigger).
Scientists already got complete blueprint of human genome and many genetically linked diseases have been identified.
However the laws and ethics prevent them.
e.g If we create 10 feet tall man but that accidently turned into spine fracture, what should we do with the unexpected outcome? We cant get rid of him and throw away.
We can create a genetically bad disease free baby. But the scientist was put in jail.
https://www.sciencemag.org/news/2019/12/chinese-scientist-who-produced-genetically-altered-babies-sentenced-3-years-jail

Humans will always be smarter than robots. WE created them!
That was true in old traditional engine.
It is not true anymore to Neural Networks.
For human knowledge, we can see human performance in bullet / hyper bullet/ ultrabullet where there is literally no time to calculate.
For chess engine knowlege, we can test with Depth 1 ( testing engine intrincic knowledge without millions of calculation).
Stockfish depth 1 is approx -1000 to 1200 elo.
Leela depth 1 is approx -2200 to 2300 elo.
Leela is way smarter than all human and she can play even NM level without a single calculation. Meanwhile, for estimate of your elo level of knowledge without calculation, take -700 to -1000 from your current elo.
IMO faster time controls only benefit AB engine even more given they have a tiny delay in their outputs and can avoid the vast majority of tactical blunders in tenths of a second, humans cannot do that consistently, hell, even Hikaru complains about relatively simple blunders in 5 min chess.

In some closed and boring positions, human chess players still play better than top chess programs.
For example in this game, Nakamura understood the standard anti-computer strategy and got a nice, closed position that modern chess engines still can't get right.
It pretty much depends on the opening they get. Sure Magnus Carlsen can definitely hold Stockfish eventually in certain simple, dry positions (the Berlin endgame for example) especially if the computer isn't blitzing out the most challenging moves properly/using sharp opening books/tablebases. GM Andrew Tang beat Leela once after getting a relatively simple position out of the opening .
That was 12 years ago, SF could beat the **** out of both Rybka and Hikaru even in such positions (that is if SF even allowed the position to become that closed).

With equal time and materials to start, can SGMs last 50 moves against Stockfish 11? U1800 players can't last 30 moves.
I rephrase my question. At what maximum Depth can SGMs beat Stockfish 11? 5 ply?
That is meaningless, the way ply counts are calculated varies across different engines and depends on how you define a searched ply (note, engines never search all possible lines up to x ply because that's dumb and in many cases, impossible).
I can already beat my computer at kick-boxing.