|Arctic sea ice extent as of 8/16/12|
It looks like it probably will. It's currently at 5.04 million square-kilometers (Mkm2), while 2007's low was 4.25 Mkm2. From this point in the year, the average remaining melt to minimum over the last decade is 0.92 Mkm2, with a standard deviation of 0.22 Mkm2.
So to break the low, this year only needs to be more than -0.60 standard deviations above the mean. If you assume the annual remaining melts are distributed normally, the probability of that happening is 72%.
Moreover, this probability has been growing over the last few days by about 2 percentage points a day. Like all such calculations, it's moderated by the fact that the Arctic is a place that usually defies simple-minded examinations like this one (except that over time, ice will always melt on a planet that is out of energy balance).
(Update, 8/17 a.m.: Overnight numbers have raised this probability to 82%.)
Why does this matter? Jennifer Francis of Rutgers gave a good talk about it: