Future Decoded is Microsoft’s annual two-day conference aimed at business strategy leaders and IT Pros. It’s been held in the UK every autumn since 2014 at ExCeL London. I had the pleasure of returning for the fifth time this year. What I love about this conference is the way it looks ahead at the big themes in technology and how they can be used to help organisations achieve more. Usually there is a business-focussed day and a technical-focussed day, however, this year it was primarily targeted at business leaders and decision-makers. I think that is down to Microsoft establishing their “Ignite: The Tour” event, which brings the more technical content to the UK each Spring.

AI showcase on the expo floor

The AI showcase had loads of great interactive examples of how artificial intelligence is transforming the world we live in

Here’s my wrap up of my visit on Day 2

 

 

AI is the new normal

My big takeaway from the whole event was Microsoft’s big push towards making Artificial Intelligence (AI) and Machine Learning (ML) part of business-as-usual. This was the main theme of the morning’s keynotes and most sessions throughout the day. There was a fascinating statistic that although 84% of their customers believe AI gives them a competitive advantage, only 14% have implemented AI. This is a big gap and reveals that businesses who have already implemented AI into their services are leaping ahead of their competitors. In previous years the keynotes have been focussed around how an Azure service X helps to do Y or trying to help people understand the importance of moving to the cloud. It was good to see Microsoft move this on, I’m assuming this means people have got the message about how great Azure is!

Mitra Azizirad, Corporate Vice President of Microsoft AI, took some time to explain what “human parity” means when it comes to AI. I’ve never really thought about it too much, surely human parity means AI is about as good as a human at certain tasks. What was then demoed on stage was the embarrassing reality that not all humans are equal and “human parity” is more comparable to experts or almost superhuman levels of understanding! The first example was the “switchboard” test;  The audience was played some audio of someone mumbling on a telephone. I didn’t have a clue what the person said, and even when presented with 3 possible options I still managed to pick the wrong one. Not an issue for the AI-enhanced program that picked it up without any issues (an industry milestone word error rate of 5.1%).

Slide detailing AI innovation from 2016 to now

Mitra Azizirad showing how AI has evolved over in recent years

The second demo was comparing a person talking, with a digitally synthesized voice, and we were challenged to guess which one was the robot. I was fooled again and incorrectly guessed the real woman was a robot (I hope she doesn’t get too offended!). They were both great examples of how AI is ready to take centre stage and even though they may not be at 100% perfection they could be some considerable margin above the average real-life person or manually coded software. This also brings up the aspect of trust and responsible AI. Microsoft takes this very seriously and has released guidelines on how to design a responsible bot.

All that being said, AI can seem out of reach for the average organisation. One imagines a team of highly trained data scientists in white lab coats running simulation after simulation. To show just how accessible it is, Microsoft’s David Carmona – GM of AI Marketing, took to the stage to show how everybody is able to take advantage of AI to help themselves and others – no degree needed.

The first was a demo of PowerPoint’s new Presenter Coach feature. This is probably deserving of its own blog post to go into more detail. David showed that this isn’t just a niche feature for people who do lots of public speaking but something anyone can take advantage of with the click of a button. The clever bit was that he wasn’t even showing a presentation, he just left the Coach running in the background while he did a live demo of something else. When he finished he went back and stopped the Coach and was presented with a dashboard showing him how he could improve it for next time. It reported on standard rehearsal features like total time spent but then shows visualisations of your pace alongside your originality, any sensitive phrases you may have used and filler words (like umm, err, obviously). I’d encourage you to give it a go yourself. Why not try turning it on in the background the next time you are on call just to see how it works (currently this is in preview in the online version of PowerPoint and may not be showing up in your tenant yet).

PowerPoint Rehearsal report live demo

David Carmona rehearsing with PowerPoint Presenter Coach live on stage

That was Microsoft’s example of how they are infusing their own software with AI for our benefit. The next example was how we can use Microsoft Azure to infuse the offerings we give to our customers/staff/citizens with a demo of a city council’s website. How many times have you got lost trying to find the right page or form on your local council’s website? It’s a common problem if your site has to cater to such a wide range of services and consumers. Wouldn’t it be easier if you could just ask for what you want and be taken to the right page? That’s exactly what you can do with bot technology and Azure Web Apps. Instead of a visitor needing to use the correct search terms or navigate a convoluted menu system, why not let people just ask the website? You may think that’s not particularly impressive now we are all used to digital assistants like Alexa and Siri but what happened next took it a step further. Not only did it find the right information but it also presented him with the correct form he needed to fill out. It then guided him through filling out the form in a very natural and user-friendly way.  

How about if you don’t have a website but you do have a phone line? That was the use case for the next demo. I’ve seen a lot of bot demos before but this seemed much more mature at the way it handled the conversation. David could speak naturally to the bot (asking for updates on his next bus) and it could have a conversation with him in a much more friendly way than a traditional “push 1 for times” telephone system. A nice thing about talking to a bot is that it’s not rude to hang up as soon as you have your answer :D. The smartest thing about this was that the audio bot wasn’t coded by a developer, it was made by a normal user of Microsoft Dynamics using the Virtual Agent app. This was a really quick drag and drop process to similar to Microsoft Flow, that anyone can use to give some structure to the bot and let the AI integration take care of the rest. David then showed how an employee could have an AI-enhanced dashboard by mixing Power BI with a conversational bot. The anomaly detection service showed a delay in the council’s transit system, sentiment analysis showed unhappy citizen’s on social media, key influencers drew out the most likely causes.

Demo slide of using conversational bot with Power BI

Infusing AI into an employee dashboard

So how do you go from an AI laggard to an AI Leader? I was pleased to see Microsoft have created an AI Business School as part of their excellent free Microsoft Learn platform. There are learning paths for defining your organisations AI strategy, introductions to AI for business leaders, how to enable an AI-ready culture, and how to develop the guidelines for Responsible AI. This looks like a fascinating wealth of content in there and definitely something I will be looking into further. Check it out and sign up here https://aka.ms/aibs

Throughout the rest of the day I went to a great mix of sessions:

  • What’s new in Windows Virtual Desktop (now Generally Available)
  • Updates to Microsoft’s Cloud Adoption Framework
  • Dreaming Starts with Innovation
  • A World View of AI

They are all worthy of their own blogs so I won’t go into detail now.

My last stop was to the AI Juice Bar. This was one of the showcase stands on the expo floor. You stand in front of a camera and get shown a series of photographs (like sunsets and mountain climbing). The service then assesses your mood based on your reactions and recommends a smoothie. I was recommended “Pick-Me-Up”, which I think means that my fully-packed day had tired me out 😁

Smoothie outside the AI juice bar

Letting my mood choose my juice

I’d encourage you to take a look at the keynote videos and explore the session recordings and presentations that have been made available at futuredecoded.microsoft.com. Share it with your colleagues as there is plenty of content available for everyone in your organisation to find something informative or inspiring.

This blog first appeared on Thom’s Headspace | You can find Thom on twitter @thommck