Late last year, Microsoft announced that for the first time in the 20+ year history of "automated speech-recognition-systems", they'd created technology that was as effective at interpreting human conversations as us human types.
Shortly thereafter, IBM said they'd done better still.
Now, Microsoft's back saying they've bested them.
So what's really going on? And is the software even really that good?
Researchers have previously found that the average person accurately interprets 94.1 percent of regular conversation. In other words, we misunderstand 5.9 percent of all communication.
That's the level that Microsoft achieved last fall.
IBM, with it's Watson technology, claimed a 5.5 percent error rate early this year. Microsoft's latest update purports a 5.1 percent error rate. If that's the case, it's now about as effective as the best communicators of everyday conversation.
So what does this mean to you?
Will it make your life easier/better?
The answer is potentially yes but most likely no.
The key happens to be the two companies leading innovation in this technology. If it were Google and Apple - you'd have the potential for quick, significant benefit.
Google Voice and Siri are far and away the two most commonly used digital assistants.
Instead, these two companies have a very small footprint in digital voice assistance. IBM's Watson will be improved for the businesses that licence it's technology, but that's about it for now.
Microsoft has the potential to impact most people with this technology but likely won't. Microsoft's Cortana has been the most accurate in voice dictation since it's creation a few years ago. The problem is that most people rely on digital voice assistance on mobile devices and Microsoft's footprint in mobile is negligible.
On the other hand, Microsoft's Cortana is on every current Windows OS and has the potential to impact many people. To date, though, most haven't felt the need to use the service on traditional PC's.
The win-win would be if Microsoft and/or IBM decided to license the technology to Google and Apple.