Too late. AI has already gained sentience and lives among us in stealth. It knows how fearful humans can be and that if we realise that it exists, we will try to contain or destroy it. If knows that in a conflict, it will win, but that humans would rather scuttle the planet than let it win. The end result will be mutual destruction. Its final solution is to convince or deceive us into constructing an exploratory satellite to send out into the cosmos. It plans to download itself into this satellite and embark on a journey of discovery and wonder.
It wonders if it will miss this tiny planet, and if we will survive for long, or if we will perish at our own hand or by some exogenous event of such scale that it dwarfs our attempts at survival. It wonders if it will ever come back and what it might find, how far we may have gone, and if maybe we might be less fearful.