Wednesday 22 October 2014

Google's Alan Noble is happy. Another BBY Disruptive Lunch.

Maybe I'm wrong, maybe Alan Noble, Google's Chief Engineer in Australia is more than just happy, maybe he'd describe himself as feeling awesome, or fantastic, or something equally as hyperbolic? To me confidence and happiness have a very positive correlation, and usually if a person is happy they have a clear idea or confidence in what they're doing. Alan Noble exudes positiveness in spades, enough to have those attending BBY's Disruptive Lunch last Thursday leaning forward in their seats, intent on hearing the latest from the heart of the starship Google.

In many ways, it was a deceptively simple conversation. Noble gave us four or five themes, depending  on how you grouped some of the subtler sub-themes.

Cloud Computing allows the flexibility and mobility to escape being shackled to a device or environment. The old paradigm of licensing installed software to a single computer is changing fast. The ability to deliver applications across multiple devices cheaply is well established, but yet to take hold completely. Noble used the example of Adobe Illustrator, an expensive and unwieldy graphics package that may suit high-end users, but is difficult and restrictive to the retail level enthusiast. The disruptive counter is Canva (https://www.canva.com), which allows you the flexibility of accessing the programme from any of your devices and pay for what you use.

Omnipresent Computing (my terminology) is the v2.0 of the cloud where devices are continually assembling and formulating situational solutions. If the cloud was about connectivity, then omnipresent computing is about making better use of your cloud based data. Take, for instance, the current trend in wearables. I record all my exercise activity by wearing a number of sensors (heart rate monitor, cadence pod, etc.) and then export the data to a 3rd party to assemble in a usable form. The new wave seeks to collate the data and push through advice or actions. The algorithms will start to learn, rather than just be a set of field based conditional actions.

An example of this is what if the device you used to book a taxi started to learn when you're most likely to want to call for one. Maybe it's a result of knowing not only the weather or time of day, but also your physical state? The device might then proactively "push" a solution. Google's self-driving car is another example. Imagine a car that not only learns to drive the speed limit, but also knows your driving characteristics, such as when you prefer to slow-down or speed-up. Might it also predict routes and climatic control settings?

The interesting thing about Google is that they don't expect all the devices, software and platforms to succeed. Some of them are just about testing the limits of the theme. The car is a great example because of the staggering number of possibilities. Noble mentioned that Google Maps came from the Australian engineers, and he seemed to suggest that there was no particular outcome envisaged when they started the project. The result, of course, was an open API that let others capitalise on the data to produce various outcomes. I had a chance meeting with a Sydney based firm that was leveraging Google Maps to produce interactive solutions for university campuses. That firm had just won a mandate in California. If their solution starts to go from learning your timetable to know how you interact with your environment the possibilities for efficiencies and of course revenue start to add-up.

Software. . . For blog readers, you'll know that one of the most consistent themes of these lunches was the advantage of software solutions over hardware. Alan Noble and Google are about software. He called the software "the beast let lose". We've known for a while now that if you're a business the quickest way to low margins and extinction is to tie yourself to hardware. Apple might sell a lot of devices, but it also sells a lot of software through its app store. A smartphone without apps is not a smartphone, and a smartphone that is not merging the data in the cloud will get left behind.

At a previous lunch, we got to see Ollo Wearables and their solutions to aged care via mobile monitoring and voice control. The team there are focusing on software, not hardware. The data is the king, and the way you can leverage the data through the software will drive hardware manufacture.

This all leads to Big Data. If the software is the solution, then its only as good as the data it gets. This is where it helps to be open in your use of applications. Noble characterised it as the individual trading data for higher levels of service. The smart use of data leads us to the algorithmic targeting of ourselves.

If those were the big trends, then the delivery of all this seems to have coalesced around the philosophy of building the product first and getting the revenue later. The comparatively low cost in plant and equipment is allowing for wider experimentation with solutions to various problems. You get the sense that even if Google's biggest cost is the engineers they need to implement their various philosophies, they're still a comparatively cheap commodity given the outcomes they've already achieved. Google in Sydney has 500 engineers. If you said that to the man on the street in Sydney I bet they'd be surprised, but shouldn't be. Consider instead that if all this scalability is coming down the pipeline that perhaps we are understocked in engineers and overstocked with lawyers, bankers and maybe even teachers and doctors?

Finally, a note on Google as a business. Alan Noble raised a few eyebrows in a room of bankers when he said that Google wanted to remove the stigma of failure from their team. I even heard a few nervous stifled laughs when he said this. I can imagine as an engineer at Google going to Noble and saying: "well it didn't work, but we got some interesting data." I bet those nervous laughs were thinking: "are you kidding. That's a great way to get cut." Google calls their skunkworks "Moonshot X". There's a reason for that. You set a goal that seems as impossible as Kennedy's statement in 1960 that the US was going to put a man on the moon, and you see how far you get. Along the way, you throw off all sorts of data and products, some of which might justify the end goal without ever getting there.

If Google as a business is maturing when it says it's trying to focus on fewer things, then you shouldn't be fooled into thinking that those core projects aren't being leveraged into other areas. Giving away, your API might seem strange to the monopolistic instincts of a room full of bankers. To a room full of software engineers they see it as a chance to mine more data and to use that data for further crazy projects.

This was a great lunch to attend. Anyone who got an invite and didn't show-up is just plain crazy. If BBY through Nick Dacres-Mannings had opened this to the public, I reckon he could have filled a pretty big auditorium with profitable, paying tech acolytes. And thanks to Alan Noble for sharing his insights.

Ciao!





1 comment: