Posts Tagged ‘Cloud’

Voice Recognition: NUI’s Unsung Hero

Wednesday, January 11th, 2012

I recently got asked to provide an opinion on “voice recognition”, in particular around our philosophy towards it and how we’ve implemented it across the stack.  If you can stomach it, you can see how it turned out (let’s put it this way, it opens with a comparison to the Hoff’s “Knight Rider”) and it kind of goes downhill from there but regardless, in doing the research, I learnt some really interesting things along the way that I thought I’d share here.

soundwave2First off, let’s start by asking how many of you know how speech recognition works these days?  Well I thought I did, but it turns out I didn’t.  Unlike the early approach, where you used to have to “train” the computer to understand you by spending hours and hours reading to your computer (which always kind of defeated the object to me), today, speech recognition works pretty much the same way they teach kids to speak/read, using phonemes, digraphs and trigraphs. The computer simply tries to recognise the shapes and patterns of the words being spoken, then using some clever logic and obviously an algorithm or two, performs some contextual analysis (makes a guess) on what is the most probable sentence or command you might be saying.

In the early days of speech recognition, the heavy lifting required was all about the listening and conversion from analogue to digital, today it’s in the algorithmic analysis on what it is most likely that you are saying.  This subtle shift has opened up probably the most significant advance in voice recognition in the last twenty years, the concept of voice recognition as a “cloud” service.

A year or so ago, I opened a CIO event for Steve Ballmer, given I was on stage first, I got a front row seat at the event and watched Ballmer up close and personal as he proceeded to tell me, and the amassed CIO’s from our 200 largest customers, that the Kinect was in fact a “cloud device”.  At the time I remember thinking, “bloody hell Steve, even for you that’s a bit of a stretch isn’t it?”.  I filed it away under “Things CEO’s say when there’s no real news” and forgot about it, until now that is when I finally realised what he meant.

Basically, because with a connected device (like Kinect), the analysis of your movements and the processing for voice recognition can now also be done in the cloud. We now have the option (with the consumer’s appropriate permission) to use those events to provide a service that continuously learns and improves.  This ultimately means that the voice recognition service you use today is actually different (and minutely inferior) to the same service that you’ll use tomorrow.   This is incredibly powerful and also shows you that the “final mile” of getting voice recognition right lies more now with the algorithm that figures out what you’re mostly likely to be saying than it does with the actual recognition of the sounds.  MSR have a number of projects underway around this (my current favourite being the MSR’s Sentence Completion Challenge), not to mention our own development around how this might apply within search.

Those of you that have been following these ramblings in the past will know I’m slightly sceptical of voice recognition, thinking that it is technology’s consistent wayward child, full of potential, yet unruly, unpredictable and relentlessly under-achieved.  I’m not saying my view has changed overnight on this, but am certainly more inclined to think it will happen, based on this single, crucial point.

Kinect too provides its own clue that we’re a lot closer than we previously thought to making voice recognition a reality, not just in the fact that it uses voice recognition as a primary mode of (natural) interaction but more in how it tries to deal with the other end of the voice recognition problem – just how do you hear _anything_ when you are sat on top of the loudest source of noise in the room (the TV) when someone 10 feet away is trying to talk to you in the middle of a movie (or the final level on Sonic Generations, sat next to a screaming 6 year old who’s entire opinion of your success as a father rests on your ability to defeat the final “boss” ).  If you have a few minutes and are interested, this is a wonderful article that talks specifically about that challenge and how we employ the use of an array of 4 microphones to try and solve the problem.  There’s still more work to be done here, but it’s a great start in what is actually an incredibly complex problem  – think about it, if I can’t even hear my wife in the middle of a game of Halo or an episode of Star Trek (original series of course) how the hell is Kinect going to hear? (Oh, I’ve just been informed by her that apparently that particular issue is actually not a technical problem… #awkward).

So these two subtle technological differences in our approach are going to make all the difference in voice recognition becoming a reality as part of a much more natural way of interacting with technology.  Once that happens, we move into the really interesting part of the problem – our expectations of what we can do with it.

expectOur kids are a great way of understanding just how much of Pandora’s box getting into voice recognition (and other more natural forms of interaction) will be and I suspect that ultimately, our greatest challenge will be living up to the expectation of what is possible across all the forms of technical interaction we have, NUI parity across devices if you like.  My son’s expectation (quite reasonably) is that if he can talk to his xBox, then he should be able to talk to any other device and furthermore, if he can ask it to play movies and navigate to games why can’t it do other things?  I was sitting doing my research with him the night before my interview on all of this, and we were playing together at getting the voice recognition to work.  He asked the xBox play his movie, he told Sonic which level to play on Kinect FreeRiders then he paused, looked at me and then back at the TV, cracked a cheeky smile and said, “Xbox, do my homework…”.

Technological Change–Above and below the water line

Wednesday, October 13th, 2010

In the final instalment of our systemic view of the changes that surround us we com to the area we talk most about, but perhaps understand the least – how technology has changed around us.

When I started my career (and I suspect it is the same for many of you), the only place I would see a personal computer was in a place of work or a place of study. Think about how different that world is now. We are surrounded by technology, much of it has become so pervasive to our everyday lives that it has started to become invisible.

When was the last time you thought about how the internal combustion engine actually works? Apart from a few petrol heads which are undoubtedly reading this, what do you do when you get in your car? Do you sit and think, <nerd voice> well, turning this key activates the fuel pump which even as I sit here is preparing the correct amount of fuel to be compressed in the cylinders and ignited at precisely the right moment, the resulting explosion creating sufficient force to drive a powertrain supplying the correct amount of longitudinal force to each of the driving wheels </nerd voice>. Of course you don’t, you get in turn the key and crack on with getting to your destination.

TechChangesmIT is becoming no different. Although the way in which we use it becomes increasingly sophisticated, we care less (or we should care less) about the specifics of what makes it work. This is a good thing. In my book, a minute spent thinking about the tool is a minute wasted as it should have been spent thinking about the task.

I like to think of it as a water line that we continue to push up as we are able to effectively “commoditise” the core elements of technology. Above the water we see the graceful, pretty technology that helps us be productive. Below the water, we know there is a complex eco-system that drives it, but we don’t necessarily need to understand every intricacy of what makes it work.

Increasingly, understanding and using this commoditisation will be the difference between success and failure.

The Efficient Enterprise in 2010

Tuesday, February 23rd, 2010

I’ve just come from a really fascinating roundtable discussion (sponsored by Dell and hosted by Bryan Glick from Computer Weekly) about the "Efficient Enterprise in 2010". The meeting was well attended by a bunch of Enterprise customers and Partners, with the context for the discussion being a presentation from Robin Johnson, Dell’s Global CIO.

There were a number of really compelling things that came out from both the presentation and the ensuing discussion:

  1. Opportunity Cost of IT Savings Understand the Opportunity Cost of any savings you make.
    OK, so you lot know I already get that one, but wow, Dell apparently are able to plough around 50% of their savings back into strategic IT (and when you’re making a $160m saving p/a, that’s a big deal). Read on to find out how they get away with that -especially at a time when most CFO’s want every penny they can get – and then some.
  2. Use the Time:Cost ratio as the pivotal argument, not simply Cost savings alone.
    Robin (and the group) talked about the difference in motivating the "business" when you factor in the time to market for IT solutions rather than simply talking about cost savings alone. It sounds simple when you say it like that, but it’s a hard won position with many CFO’s/Steering Boards. If people understand the difference in time to market that more complex IT makes, it makes it easier for them to support you in making it simpler.
  3. Pursue "Ruthless Standardisation"
    Driving a standards based architecture is a pretty tall challenge, no point in doing it then if you’re only going to go halfway. It’s tough, but if you’ve done the above, you can make it happen. Dell have only _2_ images for their 22,000 server estate. That’s pretty ruthless, but it enables them to do a lot more.
  4. Create a path of least resistance
    The Dell guys talk about the "Happy Path" vs the "Unhappy Path" when it comes to IT Architecture and solutions design. Follow the "happy path" (i.e. use standard tools/architecture etc) and you will get your solution in place more quickly and more cost effectively. It is possible to walk the "unhappy path" but it’s hard work so only those that are committed take it.
  5. "Good enough" is good enough
    It was in fact, the great Dash (from Disney’s Incredibles – see how I spare you no cultural expense on this blog ;-) that said (and I paraphrase) "When everyone is special, it actually means no-one is". Nowhere is this more true than in the internal IT vs Business debate. The more special we allow different groups/departments to be unique and special the more expensive their IT solution. This recession will force organisations and departments to come to terms with this (I hope)
  6. Rigidly define flexibility
    Oxymoronic at first blush, but it simply means, leave a little wiggle room, so people still feel empowered and part of the solution. Avoid "doing things" to people, collaborate with them instead.
  7. The Consumerisation of IT Finally (and another of my favourite topics) be cognisant of the effects of "Consumerisation"
    Robin talked about the "Sunday Night/Monday Morning" concept, whereby people have a great IT experience on Sunday night as they catch up on personal tasks on-line, then go into work the following morning to receive a comparatively poorer experience. This isn’t about embracing the millenials, but about providing a range of service that suits a range of generational stereotypes.

Although the discussion was mostly business focussed, there were a couple of key technological points that I felt we worth calling out:

  1. Power consumption is the new gold
    Based on the granularity of their server provisioning approach (smallest unit of MIP "currency" is a 2U box), Dell reckon that it is now power consumption that drives their hardware refresh cycle. Robin currently reckons that a 3 year refresh cycle will provide sufficient financial savings in power consumption alone to pay for the refresh.
  2. Virtualisation alone is not enough
    Although it took a record breaking 60 minutes into the discussion before anyone mentioned the "c" word (Cloud, that is), what was clear that a big part of Dell’s success in the rationalisation of their data centres was the automation of the server provisioning. This is a topic that we’re beginning to see again and again, a virtual server is still a server, it still needs to be provisioned and patched. You only get the big savings, when you can automate that process sufficiently (and model it so you know what you’re doing is right).

Cloud Computing – 2010

Friday, January 8th, 2010

cloudbin Just came across a good article on the future of Cloud Computing for 2010 – it’s a great summary and indicator of how fast things are progressing, you can read the full article here, but in summary, the following (in my opinion) are the three key areas to watch:

The year of the platform – after years of working through Infrastructure as a Service (IAAS) and Software as a Service (SAAS)the coming of Platform as a Service (PAAS) marks the beginning of the realisation of the full benefit that the cloud approach has to offer.

Here Come Private Clouds – specially pertinent in the Public Sector, the information assurance requirements of governments and nations are driving an approach that sees the adoption of “national cloud” infrastructure.  This will challenge many of the cloud providers as issues around data sovereignty and legislation such as the US Patriot Act will force them to offer geographically local solutions (which is kind of counter intuitive to the purists architectural vision for cloud computing).

Security – What a surprise, the growth of cloud computing has lead to a new frontier to defend and secure.  whether it’s new concerns like Differential Privacy, or old “friends” moving to new platforms (the first cloud based bot-nets have recently been discovered).

Differential Privacy

Friday, October 9th, 2009

PrivacyEarlier this week I blogged about the growing evidence of governments opening up their public data at both a national and local level. While this in itself represents a great leap forward it brings with it a new set of challenges the we will need to address. One in particular stands out and it is around the evolution of some of the very real challenges we’re going to face around Privacy in a Web/Gov 2.0 world.

Earlier this month I was chatting to Stuart Aston (one of our security advisors – you know the type, smarter than your average bear and very switched on to the evolution of the security principles we will face in an increasingly connected world) and he introduced me to the concept of “Differential Privacy“. He left me with a few white papers and a smile and a few hours later, with my head pounding and eyes bleeding (trust me you want to try and read this stuff) I finally got my head around the concept and what it’s going to mean to us as citizens.

Differential privacy is essentially, the ability to make very specific conclusions (with incredible accuracy) about the identity of an individual when provided with two disparate sets of anonymised data on a similar topic.

The example given uses NetFlix’s recent competition to improve their recommendation system as the backdrop…

DiffPriv

NetFlix published an anonymised data set of around 500,000 records in order to help developers come up with a solution to improve their recommendation system. Some bright sparks took this data and a similar export from the IMDB and by applying some fairly hairy maths, they were able to identify specific individuals with a shocking 96% accuracy rate.

This is mind blowing, not just because of the maths involved, but because of what it means in a world of growing public data, the old bastions of Privacy that we have relied upon thus far may no longer be enough.

Governments and organisations are going to need to take this seriously as it will present some difficult challenges about liability and the duty of care to keep their citizens/customers identity and data private.

In particular, think about the duty of care element. As an organisation, you have a legal requirement to look after the privacy of the data you hold on an individual or organisation – with differential privacy, how far does this duty of care extend? If you keep your data anonymised but others can compromise that privacy (albeit with hairy maths and more public data) who is actually liable or legally responsible for the breach?

There are some tough answers to be found here and undoubtedly some more legislation will be required – in the meantime though, it’s a concept we need to understand more so we can build appropriate responses that don’t restrict the overall movement towards making public data more readily accessible . We cannot afford to let this (and other similar issues) stop the democratisation of data, but we do need to go into this with our eyes open.

Cloud Architect Forum – Slides Now Available

Tuesday, October 6th, 2009

Matt‘s just posted all the slides from the Cloud Architect forum on September 25th.

There’s some great content here and some good food for thought (especially from Neil Ward-Dutton – who presented the best high level summary of “just what exactly _is_ the cloud” that I’ve seen to date).

I’ll be turning my session into this month’s podcast so keep a look out for that.

UK Broadband not fit for purpose

Thursday, October 1st, 2009

Network DownA recent Cisco report shows the UK ranking 25th out of 66 countries in regards to quality of their networks

OK, so maybe not fit for purpose is a bit strong, but it does go to show that our ability to achieve some of the things we’ve talked about (especially cloud based solutions) is going to face some pretty difficult challenges.

Bandwidth is the unspoken barrier to the cloud, we assume it to be there but in reality it isn’t consistent enough (yet) for us to rely on it providing the necessary connection to all the great things we want to achieve.

Application architects and solution providers will need to think creatively about how we build applications that take this into account.

Cloud Architect Forum

Friday, August 28th, 2009

dataMatt recently invited me to provide a session for the Cloud Architect Session he is going to run in London on September 25th.

He’s already got some great speakers lined up – Neil Ward-Dutton from MW Advisors, Simon Guest, Senior Director, DPE (Microsoft) and Andrew Jones, Practice Director from CSC.

Attendance is free (limited to first come, first served) and you can register here.

FWIW – this was the overview I gave Matt for my talk:

“Cloud Computing is being positioned as the future of the IT Industry, yet we collectively seem to be unsure on what it actually is and, more importantly, how to get from here to there. This session will attempt to reconnect us all with the longer term reasons on why the cloud is not just important but fundamental to the kind of transformation we are looking technology to provide the way we live, work and play.”

All I need to do now is come up with the slides to match…