How Old Robot – Well played

How Old example
Example of How Old Robot doing its thing

So, a few days ago, the internet had a new toy: How Old Robot – a very simple website where you can upload a photograph and it will guess your age and gender.  For many people the guess was about right, but there were some howlers, with very similar images being uploaded and giving age results differing by (several) decades!

The site doesn’t hide the fact that it’s a learning tool based on Microsoft’s facial recognition technology and is built on the Azure platform as an example of how quickly it is to build and deploy sites using Azure.  What started off as a quick demo from the Build conference soon became viral, with people all over the world loading their photos into the app and sharing the results on social media.  This is exactly what Microsoft wanted and they’ve been oh so clever with this and here’s why.

Firstly, this immediately became the easiest advertising of their technology they’ve had in a while.  Developers all around the world are looking at both the Azure stack for rapid deployment and the new facial recognition technology (despite it’s flaws!).  This site is a wonderful example of what can be done without having to write tools from scratch.

Secondly, millions of images have been uploaded in a very short space of time.  One of the biggest hurdles with any deep learning project is finding the data to train the system initially and then continue to fine tune with as many new examples as possible.  While the Microsoft facial recognition technology already had been trained to recognise faces in images, gender and age to some extent, this will have been on whatever data set the developers had access to – how better to train their system than encourage worldwide users to upload their own images in all their variety.  There’s a reason the app apologises: “Sorry if we didn’t quite get the age and gender right – we are still improving this feature” and they are relying on people uploading images to do this improvement for them.

Finally, I would bet that many people didn’t read the small print – a few others have pointed this out:

However, by posting, uploading, inputting, providing, or submitting your Submission, you are granting Microsoft, its affiliated companies, and necessary sublicensees permission to use your Submission in connection with the operation of their Internet businesses (including, without limitation, all Microsoft services), including, without limitation, the license rights to: copy, distribute, transmit, publicly display, publicly perform, reproduce, edit, translate, and reformat your Submission; to publish your name in connection with your Submission; and to sublicense such rights to any supplier of the Website Services.

So any images you upload they can use however they see fit.  There is a note on the site saying they won’t keep the photos, but they say nothing about any meta data or any processed version of the images that will help them use everything from your image without actually keeping the photo itself.  While it’s unlikely that your holiday snaps will make it into Microsoft’s next advertising feature it is a gentle reminder to a) read terms and conditions and b) if you use free services you are paying for the service in non-monetary ways.

On the whole, well played Microsoft – they’ve knocked together a very simple app to showcase some of their technology and have released it to the world to let the community do the hard work to train their algorithms while at the same time harvesting data from millions of images for this and any other purpose they see fit. Very well played.

janet
Dr Janet is a Molecular Biochemistry graduate from Oxford University with a doctorate in Computational Neuroscience from Sussex. I’m currently studying for a third degree in Mathematics with Open University.

During the day, and sometimes out of hours, I work as a Chief Science Officer. You can read all about that on my LinkedIn page.

Published by

janet

Dr Janet is a Molecular Biochemistry graduate from Oxford University with a doctorate in Computational Neuroscience from Sussex. I’m currently studying for a third degree in Mathematics with Open University. During the day, and sometimes out of hours, I work as a Chief Science Officer. You can read all about that on my LinkedIn page.