On your first point you should read the question of nationalities which Lenin wrote shortly before his death. He clearly wanted to take down the tsarist apparatus after all the existential threats to the Soviet Union were gone.
Where did Luxembourg say Lenin was trying to recreate the tsarist empire? She was critical of the Bolsheviks authoritarianism but If anything she was also critical of the Bolsheviks limited allowance for nationalism and would’ve suppressed nationalism further, she was a strict internationalist.
If they did dismantle the state apparatus before the Nazis came what do you think would happen? The Soviet Union was barely able to turn the tide of the war with a united front and 20 years of intense, brutal industrialization. If they had dismantled the state and Russia was just a bunch of rural locally run villages in a loose confederation in 1939 the Nazis would’ve steamrolled over them and genocided the population.
There are 2 problems with not having enough diversity in training data:
The AI will be worse at depicting diversity when prompted, eg. If the AI hasn’t seen enough pictures of black people it may not be able to depict black hair properly as it doesn’t “know what it looks like”
The AI will not show as much diversity when not prompted. The AI is working off statistics so if you tell it to depict a person and most of the people it’s “seen” are white men it will almost always depict a white man because that’s statistically what a person is according to its data.
This method combats the second problem, but not the first. The first can mostly be solved by generally scaling the training data though, which is mostly what these companies have been doing. Even if only 1% of your images are of POC, if you have 1b images 10mil will be of POC which may be enough to train it. The second problem would remain unsolved though since the AI will always go with the statistically safe 99%.