In the past 10 years new and more accurate stellar neutron capture cross section measurements have changed and improved the abundance predictions of the weak s process. Among other elements in the region between iron and strontium, most of the copper abundance observed today in the solar system distribution was produced by the s process in massive stars. However, experimental data for the stellar 63Ni(n,γ?)64Ni cross section are still missing, but is strongly required for a reliable prediction of the copper abundances. 63Ni (t1/2=101.2 a) is a branching point and also a bottleneck in the weak s process flow, and behaves differently during core He and shell C burning. During core He burning the reaction flow proceeds via β?-decay to 63Cu, and a change of the 63Ni(n,γ? )64Ni cross section would have no influence. However, this behavior changes at higher temperatures and neutron densities during the shell C burning phase. Under these conditions, a significant amount of the s process nucleosynthesis flow is passing through the channel 62Ni(n,γ?) 63Ni(n,γ?)64Ni. At present only theoretical estimates are available for the 63Ni(n,γ?)64Ni cross section. The corresponding uncertainty affects the production of 63Cu in present s process nucleosynthesis calculations and propagates to the abundances of the heavier species up to A=70. So far, experimental information is also missing for the inverse 64Ni(γ?,n) channel. We have measured for the first time the 64Ni(γ?,n) 63Ni cross section and also combined for the first time successfully the photoactivation technique with subsequent Accelerator Mass Spectrometry (AMS). The activations at the ELBE facility in Dresden-Rossendorf were followed by the 63Ni/64Ni determination with AMS at the MLL accelerator laboratory in Garching. First results indicate that theoretical predictions have overestimated this cross section up to now. If this also holds for the inverse channel 63Ni(n,γ?)64Ni, more 63Ni is accumulated during the high neutron density regime of the C shell that will contribute to the final abundance of 63Cu by radiogenic decay. In this case, also a lower s process efficiency is expected for the heavier species along the neutron capture path up to the Ga-Ge region.