Controlling "memory" in a petri dish

Well, if a firing pattern is considered to be a memory. A group of cultured neurons fired in a synchronized, replicable manner with a chemical that blocks some inhibitory neurons. Entrainment towards days of unassisted firing was possible by repeated, regular dosing. Changing location would initiate a pattern for a different cluster of neurons, but wouldn't interfere with the firing pattern of the first set -- layering of memories, as it were.
Just have to get the encoding-to-silicon part down and we have a way to recreate memories, I suppose.