About the same as it is now.
Here is a more detailed answer. You ask a very good question; which relates to a dispute over how long it would take for the Earth to cool from a molten state to its present state. The question received a lot of attention around the end of the nineteenth century, when Kelvin estimated that the Earth could be only around about 25 million years old; which was far too young to encompass what was being discovered about gradual processes in geology and biology for developing the features seen on the Earth.
The solution to the dilemma was radioactivity, discovered at the turn of the century. This meant heat was being continually generated within the Earth, and calculations based on a cooling sphere did not apply.
But the data used for estimating cooling of the Earth was not a surface temperature. It was the temperature gradient within the Earth. The surface temperature is determined almost entirely by how solar radiation interacts at the surface; with the effect of geothermal heat being quite negligible. Without the nuclear reactions at the core, the whole Earth would have long ago cooled down to be something similar to the surface temperature. As it is, the interior of the Earth remains very hot, with a strong temperature gradient.
Mathematical details of Kelvin's famous argument can be found at Lord Kelvin and the Age of the Earth
, a paper used within a mathematical unit at the University of Rochester, NY, USA.
Felicitations -- sylas