The reason your logic doesn't work out is because the assumptions you put in place are counterfactual.
What is factual is that the nuclear industry has been around for many decades and has not been able to solve this problem, despite there being lots of pretty clever people involved in it.
A contributing factor to making the learning curve negative is that the security guarantees underlying the plant designs are built on what we know can go wrong. As we learn more about the problem, we also learn of more weird failure modes, complicating the design of future plants.
I've seen people handwave this away with paper designs. Paper designs are comparatively easy to make safe. Real reactors are much, much harder. If you think this is bullshit, try looking up a certain admiral Rickover and see what he has to say about this.
What is factual is that the nuclear industry has been around for many decades and has not been able to solve this problem, despite there being lots of pretty clever people involved in it.
A contributing factor to making the learning curve negative is that the security guarantees underlying the plant designs are built on what we know can go wrong. As we learn more about the problem, we also learn of more weird failure modes, complicating the design of future plants.
I've seen people handwave this away with paper designs. Paper designs are comparatively easy to make safe. Real reactors are much, much harder. If you think this is bullshit, try looking up a certain admiral Rickover and see what he has to say about this.
Meanwhile, renewables keep falling in price.