"we know not to do the thing that makes them explode" This is not the core problem. The core problem is that explosion can still happen - even if we "know not to do the thing that makes them explode". It can happen because of a natural disaster, because of war and because of other unpredictable reason. To avoid the possible explosion and consequences altogether it is better to not have them in the first place!

> we know not to do the thing that makes them explode.

We collectively know a lot of thing, but as soon as you leave the opportunity to a human an accident will happen sooner or later. We know not to start heavy machinery while there is a maintenance technician inside it, but it still happens.

This is why we have safety lockout systems, to avoid an accident even if we know better.

The RBMK has a lot of issues that make it really unsuitable for further development:

1. Because of the size of the core, it develops uneven reactivity during operation and requires attention to be paid to this.

2. The measures put in place do not entirely eliminate the chance of a power excursion.

3. The cost of a suitable containment building would make RBMK more expensive than an equivalent VVER installation. Also as far as I can tell no such containment building has ever been designed.

4. No international agency or government would allow the construction of a new RBMK. Edit: to be clear, this is not only because of the safety issues, but also because of the nuclear weapon proliferation issues associated with the design.

Overall, the cost savings of a RBMK depend on not having a large containment structure and the RBMK was on its way out as a design by time the Chernobyl accident happened. The large number of RBMKs under construction at the time was largely because of initiatives to increase the installed nuclear capacity in the Soviet Union during the mid 80s, existing RBMK sites would receive additional reactors and all but one of the new sites were to receive VVER reactors.