In todays class I learnt how to zip up a file on a mac. So unlike PC’s, mac automatically unzip your files. Which doesn’t help when trying to carry out tutorial tasks. It’s not hard but whenever I hit roadblocks and things don’t work perfectly as detailed in the steps I stop, instead of pushing through to find the answer straight away. Anyway I found the answer in record Amanda time today and can now move on with the rest of the task. So I tried the Reduction task with interesting results. After uploading the zipped files and waiting well over the suggested 3-4 minutes training time, I was able to see how the recognition program works.
Every week we are shown something new that I would love to implement on some kind of cultural heritage endeavour. If seems fairly simple using the visual recognition program and I wonder if it could be used to identify family heirlooms or museum acquisitions. You know how people take their treasured possessions to some expert on ‘Antique Road Show’ or a local expert. What about a program that people could use to identify their object at home. I guess the database would need to include millions of images and information collaborated by many experts. But it could be possible. This type of program would be particularly useful for classifying museum acquisitions especially in smaller museums where an expert may not be present. Once the object has been identified it could be presented with a list of history, display, and care instructions. 3D imaging would also help the accuracy of the recognition program.
As for recognition use in the future with all its dangers and issues, humans have always recorded the past. And like the use of redactions, decided for whatever reason to include or not include parts of those details in their retelling of their version of the past. We, as a human race have managed to mold our past as accurately as our memories and personal requests have desired. So, yes using technologies might also skew the history records, but that really is no different to what has previously been happening. So I think the danger here is not so much in the false positives a machine gives but the power we are giving to machines in writing our past.
This isn’t all doom and gloom though whatever the out come the human race will still move forward and deal with whatever catastrophe we create by our own doing and it will be just another story for the history book. If technology is going to keep growing in use and advances as it has (and of course it will) then people just need to soften up to the fact that yes they may one day be labelled a gorilla like Google’s photo app mishap, or Asian when frowning (as in one of Tim’s facial recognition examples) or any other number of titles. Besides I don’t think the new fashion styles for avoiding facial recognition are that bad. They are very futuristic and the future is where we are all heading.