There's a pretty important article in this month's Geology magazine (link is only to the abstract) which questions the geologic evidence used for the very beginnings of life. The article, reviewed by Stephen Moorbath , shows that rocks taken from Akilia, Greenland, which had been reported to show chemical traces of life, are not of the relevant type for this effort. Moorbath concludes, rather conservatively, that for the moment, the oldest certain traces of life are bacterial fossils "only" 2 billion years old. ( See Wikipedia for a timeline that needs to be reconsidered and general background.)
Everyone agrees that life has been around for a long time, but the very oldest rocks are hard to find, and frequently heavily distorted. Geologists can usually still identify heavily altered sedimentary rocks, and Australia and Greenland (in particular the Isua peninsula) have formations which are 3.8 million years old, corresponding to nearly the end of the Late Heavy Bombardment . Geologists can measure these rocks for "biogenic" (suggestive of life) traces, because living things preferentially use carbon-12 over carbon-13, whereas geological processes do not. It was therefore astonishing that graphite grains in the Akilia rocks were reported to have exactly this carbon signature, meaning that life emerged essentially right after the last big impact.
The current paper delivers a killing blow to the Akilia work. Not only is the dating of the rocks a bit problematic, but in fact the precise samples used don't have any graphite in them! These samples simply cannot have a carbon signature one way or the other. In his review, Moorbath, who has been heavily involved in work on nearby Isua, also says that those rocks have no distinctive carbon signature.
I should say that Moorbath is maybe a bit pessimistic. Although the carbon work is toast, I think the Banded Iron Formations , which must form in the presence of oxygen, at least show progressive transformation of the Earth's atmosphere. If this is taken as the product of life, then life would already be global at nearly 3 billion years ago.
These earliest dates are important because they affect surmises about the likelihood of life outside of Earth. Assuming life developed on Earth (and wasn't imported), faster occurence of life may imply less stringent requirements. That is, something which requires the conjunction of 5 sequential one-in-a-billion events would be expected,on average, to take longer than something which only requires 3 such events. We only have n=1 here, so I wouldn't hang my hat on it, and talk of averages is a bit improper. Nevertheless, if life is "only" 3 billion years old, my guess would be that it needed some time and a lot of luck to emerge here. If evidence of life is swimming around in the very oldest rocks we can find, that's really something, and maybe life really is elsewhere as well.