The earliest reports of ferromagnetism date to Thales of Miletus who lived and wrote around 600 BC. Thales noted the ability of natural magnetite to attract iron, and is said to have taken this as proof that matter itself was alive. Our theories of magnetism have evolved considerably since then: we now know that ferromagnetism arises from the interplay of the Coulomb repulsion between electrons and their fermionic statistics. However, in one sense our science has advanced only little: the vast majority of magnets, like magnetite, consist of ordered arrangements of the electron spins stabilized by the spin orbit interaction. I will describe experiments probing magnetic states based on the spontaneous alignment of electron orbitals. Such orbital ferromagnetism may be a generic phenomena, but has, to date, found its fullest expression in graphene heterostructures in which the two dimensional orbits of electrons in distinct momentum space valleys provide the underlying degree of freedom. As an elementary example I will show data from rhombohedral trilayer graphene, where band edge van Hove singularities lead to a cascade of transitions between metallic ferromagnetic states distinguished by different broken valley and spin symmetries. Adding a moire potential to the trilayer by hBN alignment allows for energy gaps at finite density when the underlying degeneracy of the Fermi surface matches the superlattice filling factor. Because orbital degrees of freedom arise directly from the band wavefunctions, they are uniquely susceptible to experimental control via materials design and new forms of magnetic control using in situ knobs. I will show examples from a variety of moiré heterostructures where magnetic moments, and the resulting quantized anomalous Hall effects, can be tuned using electric currents and the electric field effect.