This talk will give a personal view of Hamiltonian truncation for quantum field theory. This is a variational approximation of quantum field theory formulated directly in the continuum, in which the Hamiltonian of the theory is numerically diagonalized on a large but finite subspace of states.  This method can preserve symmetries that are broken by the lattice, for example chiral symmetries and supersymmetry, and therefore can be applied to theories that are otherwise inaccessible to numerical approximation. The truncated theory is a non-local effective theory, but with a systematic power counting that allows for an "improvement" program similar to the one used on the lattice. I will discuss some of the promise and challenges of Hamiltonian truncation, including the application to chiral gauge theories.