The presenter is Ross Rhodes here he explains an experiment performed: http://webcache.googleusercontent.c...-boggling&cd=1&hl=en&ct=clnk&client=firefox-a "Record the measurements at the slits, but then erase it before analyzing the results at the back wall. Suppose we take our modified double slit set up -- with electron detectors at the slits -- and still leave everything intact. And we will still keep the electron detectors at the slits turned on, so that they will be doing whatever they do to detect electrons at the slits. And we will record the count at the slits, so that we will be able to obtain the results. But, we will erase the data obtained from the electron detectors at the slits before we analyze the data from the back wall. The result upon analysis: an interference pattern at the back wall. Notice that, in this variation, the double slit experiment with detectors at the slits is completed in every respect by the time we choose to erase the recorded data. Up to that point, there is no difference in our procedure here and in our initial procedure ([pp. 15-17]), which yielded the puzzling clumping pattern. Yet, it seems that if we, in a sense, retroactively remove the electron detectors at the slits (not by going back in time to physically remove them, but only by removing the information they have gathered so that it is not available from the time of the erasure going forward into the future), we can "change" the results of a completed experiment, so far as those results are determined by a later analysis, to produce an interference pattern instead of a clumping pattern. This is mind-boggling." I think he's describing the double slit eraser and I dont think he knows what "erasing the data before anylizing it" actually entailes or means. When you "erase" data in a double slit before reading the "information" youre sticking a photon/ electron polarizing filter, youre actually altering the measuring device. Also the "information" appears to be the photon/electron interaction not a set of numerical data on a computer. WHY CANT QM USE PRESICE LANGUAGE!!!!