Heating water on a hot plate is safe, because the hottest point is at the bottom of the pot. A lot of relatively small bubbles appear there without much overheating of the water, because there is a lot of nucleation at the uneven phase boundary steel-water.

In a microwave, the hottest place is IN the water. The glass does not get heated by microwave (at least not much), and radiates off some heat to the surrounding.

Problem: In clean water, there are few good nucleation points to form bubbles, only some dust particles perhaps. So the water gets overheated rather strongly, and a first bubble that appears can grow a lot before it has cooled its surroundings down to 100°C. That one huge bubble can throw most of the water out of the glass. It boils over, violently.

Btw. a microwave does not heat its content uniformly. It forms a standing electromagnetic wave (that's not radiation, stricly speaking) in the oven, like a rope swung quickly between two people, or a guitar string. The wave pattern has knots at a distance of $0.5 c/f \approx 6\rm{cm}$ (with $f=2.45\rm{GHz}$ and speed of light $c$ which of course is a bit less *in* your chicken), where there is very little heating. That's why the microwave oven has the rotating plate, to generate *some* uniform heating. For pure water, that doesn't matter, because convection sets in anyway and distributes the heat.