HST is commonly thought of as an optical-IR imaging or UV-spectroscopy observatory. However, the advent of WFC3-IR made it possible to do slitless infrared spectroscopic surveys over an area significant for galaxy evolution studies (~0.15 deg^2). Slitless infrared spectroscopy is uniquely possible from space due to the reduced background. Redshift surveys with WFC3-IR offer probes of the astrophysics of the galaxy population at z=1-3 from line features, and the true redshift and spatial distribution of galaxies, that cannot be done with photometric surveys alone. While HST slitless spectroscopy is low spectral resolution, its high multiplex advantage makes it competitive with future ground based IR spectrographs, its flux calibration is stable, and its high spatial resolution allows measuring the spatial extent of emission lines, which only HST can do currently for large numbers of objects. A deeper slitless IR spectroscopic survey over hundreds of arcmin^2 (eg one or more GOODS fields) is one of the remaining niches for large galaxy evolution studies with HST, and would produce a sample of thousands of spectroscopically confirmed galaxies at 1<z<3 to H=25 and beyond, of great interest to a large community of investigators. Finally, although JWST multislit spectroscopy will outstrip HST in resolution and sensitivity, I believe it is critical to have a spectroscopic sample in hand before JWST flies. This applies scientifically, to be prepared for the questions we want to answer with JWST, and observationally, because JWSTs lifetime is limited and a classic problem in targeted spectroscopy has been the turn-around time for designing surveys and for deciding which classes of objects to target. This white paper is released publicly to stimulate open discussion of future large HST programs.