While this may not be a photography device, it may have an impact on those that rely on a photo for ID purposes. I recently purchased this device that uses AI technology to interpret and identify species using sound. In the 3 hrs of trial it has delivered 15 different species from 43 detections offering me a downloadable CSV file which I believe to be accurate. This is not an attempt to support the product however I feel this detection unit may be used by many to substantiate surveys, be used to support ecological reports for developers / Council etc, all based on AI interpretation. Will this also have an impact on those photographers who use their skills in this area to provide visual identification? Any thoughts or is this just not a photographical subject?

Hello Greg

There are already apps which attempt to match bird calls.  An example is the Merlin app from Cornell Lab (which I think also hosts eBird) for iPhone and Android.  This attempts to match bird calls with stored call recordings, although according to the web site this currently only operates in some countries which does not yet include Australia.  I expect that as the samples of audio available to these apps increase, the more accurate they will become.  

Most app-based bird field guides now include species by species bird call recordings, and there are stand-alone bird call apps such as Stewart Sounds.

I think that developments in call-matching technology can only help bird watchers and others involved in ornithological work with bird identification and monitoring.  I'm not sure that AI-based sound identification will have a huge or transformational impact in the behaviour and work of such people, as experienced amateurs and professionals often become adept at identifying calls across a wide range of species, using their own field observations as well as the tools already available. However, people are always finding new ways of using technology so you never know what might develop in the future.  Perhaps sound recording and matching can play a role in some types of research such as remote monitoring of rare bird activity in an area, in a similar way to trail cameras.  However in the short term at least, I doubt anybody will be happy relying totally on computer-based matching alone, without some form of person-based verification. 

Thanks
Simon