This is the most exciting Microsoft Build for me, ever, for two reasons. Firstly, this is my first Microsoft Build as a Microsoft Regional Director and it is giving me new opportunities to see how developers are empowered to have one of the most impactful jobs in the environment. Secondly, this edition of Microsoft Build is all about Artificial Intelligence for me. I never thought that my decision to do postgraduate work in AI, two decades ago, would still be my passion twenty years on. Now, it seems that the world is catching up with AI again and this time, it’s not going to disappear from organizational radars.
“The focus of innovations and investments need to shift from slogging innovations like phone and, drone type innovations to solve real problems that are killing the inhabitants of the planet.” ― Saurabh Gupta, Earth5R
Following on from Gupta’s quote, I disagree with this emphasis. Used wisely, I think drones can have a good benefit. In natural and manmade environmental disasters, drones can be positioned to survey damage, locate stranded and injured victims, and assess ongoing threats. This means that we don’t risk the safety of rescue teams and first-responders. Drones can help with tactical surveillance and suspect tracking; if they are helping you or your family members, you may well be grateful for them. I can see AI and IoT used for good, here.
I’m excited by two main things that Microsoft are doing: AI on IoT Edge, and the Open Sourcing of IoT Edge Runtime.
Cognitive Services support on IoT Edge so that you can build apps that leverage Artificial Intelligence algorithms on your IoT devices. It’s more than just hardware, though: it means you can help your devices to listen, see, speak and even interpret their environments. This includes drones but could include other things, such as industrial devices. Microsoft is providing you with Azure Machine Learning packages which allow developers to develop with devices to give you vision, text and forecasting. I think developers will be hugely empowered to do great things with Artificial Intelligence and Internet of Things technology. Microsoft are going in the right direction with Azure AI and IoT, and I think we can see AI for good, with Azure as an engine to move the world forward.
Having worked with computer vision 20 years ago, I’m excited how Microsoft are making Artificial Intelligence easy and accessible.
I think developers will be genuinely excited about developing with devices such as drones. Who wouldn’t? Custom Vision is the first Azure Cognitive Service to support edge deployment, with more coming to Azure IoT Edge over the next several months.
I worked with computer vision very early in my AI career, in a project where we looked for edges, blobs and convolution were part of my everyday vocabularly; the Laplacian of Gaussian, anyone? Twenty years ago, computer vision was extremely tough to do, and I looked at doing it 20 years ago when it was AI was considered as a route to recognise objects, particularly faces e.g. identifying shoplifters as they entered a shop. Now, Microsoft are taking it further so you could recognize things, and then take actions when something is recognized. How amazing is that?
I’ll be honest; two decades ago, computer vision was horrible and probably one of the hardest programing things that I ever did. I used C# and worked in a Computing Science department where people considered that Emacs was for softies.
I’m in awe that Microsoft are now making Artificial Intelligence so that it is accessible and easy.
What do Microsoft’s announcements mean for developers?
If you’re a graduate, or considering a career, I’d throw myself right at Artificial Intelligence in the cloud. I wish I could life my life again so I could have these opportunities as a young adult.
What does this give you, regardless of your career trajectory? Opportunity. It is a leveller to have the intelligent edge, whatever your size. It is for small organizations, large organizations, students who want to explore. These packages are not full-blown models with simple APIs the way Azure Cognitive Services are. At the other end of the spectrum, they are not raw algorithms, such as those offered in CNTK or TensorFlow. Instead, they are models that can be customized for particular applications; quicker roads to success, and that provides your opportunity to access success more quickly.
With great power, comes great responsibility (and opportunity)
As every Spiderman fan knows, with great power, comes great responsibility. We have to ensure that these technologies are empowering everyone in every industry. It’s about creating and developing trust. As quoted by Satya Nadella, Jonas formulated this imperative inter alia as follows: “Act so that the effects of your action are compatible with the permanence of genuine human life,” or, expressed in negative terms: “Act so that the effects of your action are not destructive for the future possibility of such life.” When we work with technology that offers great opportunity, we have to think about acting with great responsibility.
Great responsibility in the everyday
One example: GDPR is a sound regulation with good intentions at heart. Microsoft have been diligently working with great responsibility in mind, whilst balancing these responsibility with great opportunity. We can ask what computers can do, and what computers should do. I would take this further; we need to ask ourselves what we can do with computers. Let’s hope for good intentions and that these are enabled through ethical AI. We have the opportunity to do great things, to lift humankind forward, and increase communication. I’d personally love to see a situation where Facebook ads and fake news sites are fact-checked as you open the site, so that you can get a real view of the data when you read it. I’d love to see a proactive, artificial intelligence fact-checker that can be put beside Breitbart, for example, so that the facts aren’t as easy to ignore.
Open Sourcing of IoT Edge Runtime
Developers can have more transparency and control for their edge application since they now have the power to modify the runtime. This is a great move since it gives open source developers the opportunity to get involved with IoT Edge, springboarding from their existing expertise.
How does this affect you?
The world is speeding up and it’s not going to stop so you can hop onboard. Get learning, get reading, and join in. Satya Nadella mentioned a few books and authors during his keynote. Why not start there? Even if you don’t code, you don’t need to see the possibilities and opportunities to be the ‘glue’ that holds AI, IT, and the business together. All aboard!