Jest <15 used Jasmine1 in the background for test execution. Many developers complained about Jest being slow, and finger pointed to Jasmine. It suffered slow startup times, slow execution times and not pretty error reports. Facebook devs ignored Jasmine2 for a long time, and it looks to me that Facebook may give up support for Jest at all since ava.js hit the ground.
But surprisingly, Facebook pushed Jest very much in the last couple of weeks. Replacing Jasmine looks to me like stepping up and moving forward; become a competitive alternative to Jasmine2, Mocha, and Ava.