Does data understand the meaning of a wink? #ds13
Does data understand the meaning of a wink?
I believe that the problem we have with understanding data..... is the same fundamental problem, that we have about our views,
our independent views,
our independent views based on experience,
our refined independent views based on on-going experience ; that we suffer when we talk about any political hot potato such as the economy, bank debt, personal credit, environmental change, global warming, privacy, size of government, policing or marriage reform…… we all have unique views and our views are different from the others around us and our views are also different again from our customers views – which apparently are the only ones that count!
I contend that personal information, personal data, your data, your digital footprint and its relationship to you, your identity and your rights has the same complex mix, blends and balances that set and counter your personal views and insights. Experiences refines them and they all combine to form your own personal opinion - your view - what you want. Whilst it is safe to accept that we all have unique opinions and don’t have a common view ..... we do all fit into certain buckets or segments; I will argue that we need to get over our personal prejudices, which bucket or segment I sit in, if we want to make progress. Progress in the use and application of data!
So the question Does data understand the meaning of a wink?
I can use a sensor to determine you have winked - so let's be honest and accept that I can get the data, it might not be elegant at the monument but it will get there. 3 years ago we had had the iPad for less than 1 month, 30 days, 1000 hours. It moves fast.
I can, by applying algorithmic determination know the possible meanings of the wink. Be honest, there are only a certain number of meanings a wink can have - and it will be one of them.
Will I be able to get the context of that wink - In truth, I know this hurts, but yes. I can pick up who you are speaking with, where you are, what you just did, how you speak to this person, what the topic you are talking about means, I can get context - get over it.
But will I or data ever truly know the meaning of a wink - let’s consider this ...
Does data understand a wink?
Generally those who think about data will agree that good data, run over a good algorithm will present insights about that data set. The basis of creating value from this data set is however predicated on the assumption that your behaviours are similar to of those who behave like you in the context you are currently in. As long as I have enough data about others who are like you, within this context - I can make algorithms - digital processes to mimic you, your behaviour and predict what you will do, and what that wink could mean or imply - this body of work is behavioural economics.
Does data understand humans.
Does data help interpret or predict what we humans do based on what is a normalised behaviour for someone who is like you. Probably yes. But does data understand humans? Probably no ....- it does allow us to predict that if you’re like someone else. Therefore what we need is a new model for you and only you if we are to make progress. So how do I get more of you?
If I had more data would that help?
The key phrase here is more data, more of the same types of data that I already have.....probably no. The reason I say this is your only need one name and one address, having your name and address in 100 different silos doesn’t have any value . If I had access to new data sets then yes more data would help. More data sets could be friends, network, likes, views, shares, gossip, purchases, travel, blood pressure, heart beat, food intake, environmental conditions, music, conversations, traffic - all that yummy data that we are now collecting. Let's be careful here - this is not Big Data, the aggregation of all of us and how we behave as a population, this is Small data - all about you.
So if I have the data, what else do I need to understand a wink ?
WATSON, the IBM machine learning computer, made famous by the US TV program Jeopardy, pointed the way that computers can learn faster than we can and therefore with new data from sensors; yes..... I will know the meaning of your wink but can we determine the possible sentiments of that wink? I will get it right and your lies and excuses will not wash.
Will my machine cope with the random creativity of humanness and serendipity? probably not yet, however I would contend that the problem is not data or trying to understand the sentiment of a wink. You will have noticed I have ended up with a positive conclusion. I will know you have winked, I will know your context, I know what a wink is for and I will be able to know from your data what that wink means. Ta da...
However this is not the problem that we are facing - the problem is not about gathering data, understanding data or what data tells us - our problem is how we interface with the machines?
I believe we are fast approaching a new reality. The new reality is that the tectonic plates of what we want and what we get are smashing into each other. What we want is clear, what we get is usually dependent on someone else. In my opening I set out that we all hold different views about your data and the person next to you is as different as mine. The realities are that we will soon have a choice about data, sensors, value and engagement. Will you let the machine share your meaning of that wink?
The choices we are facing are engage, engage more, engage less, engage on someone's terms or the alternative scenario = disengage.
The famous Andy Warhol quote from 1968 which in Internet speak is 15 generations of iPads - I suggest needs to be updated to bring it in line with this fully digital age in which we live. It should no longer read " In the future everyone will be world famous for 15 minutes" but will now read. Today everyone will only be anonymous from the world for 15 minutes"
Why do I say this? 10 days ago I shared a link about Facebook and Facebooks apps. You will find it on My Digital Footprint. The blog detailed a web link, which I posted and if you follow it and are logged into Facebook, what you would see is everybody else who was using the same applications as you . If the application was candy crush you would not be a problem in knowing who is also using it, however the particular link and application I posted about was the more infamous one called "bang with friends" - The anonymous, simple, fun way to find friends who are down for the night. Quote This link is a feature of Facebook - an API - not a bug.... What the data showed me was that 128 of 980 friends I have, subscribed to this application - today it is 3! As soon as I posted the link I heard back immediately from a friend who thanked me sharing it - he had apparently tried it in a drunken fit and had forgotten. What this showed me - and I am quite a nerdy data freak - was that sometimes we don’t understand the implications of our data, clicks, likes, joins, shares, sensors and sharing
I present to you that a new reality is approaching and that the tectonic plates of what we want and what we get are already rubbed up against each other, but will we get earthquakes, volcanoes, tsunami fallout and disaster or mountains, views, achievements, scale and beauty? We have created a digital engagement platform but we don’t realise what we’re doing . Your data is telling me something you did not intend.... and there are plenty of examples of this already, including the man whose mistress was pregnant, but did not yet know and the supermarket sent the congratulations to the wife along with free diapers... opps
Our new business models depend on data. We need customers to be engaged and giving us access to their data. Putting it more starkly. No data, no model. No business. No $200bn industry in 2020, a big fat #fail.
Whilst the data we can now get tells us in incredible detail about our ourselves and our customers.
Our problem is not how much we believe is possible to obtain from data.
Our problem is not how much we can get from sensors .
Our problem is not new data sets .
Our problem is not the belief we can understand the sentiment of the wink .
Our problem is that the user interface and user experience is simply not doing a very good job . We are not bringing our customers with us . We are not explaining what we're doing . We are not sure ourselves what we are doing. Ask my wife - she tells me every day. Now there is a big dose of reality.
The reason I believe that the tectonic plates of what I want and what I get are smashing together is that I can see that we could create disengagement faster than we will create engagement . If customers disengage, no data, no model. We need customer engagement. If we create a digital social system of disengagement, customers opting out, we will sink back into the old economics of hard labour and supply demand models . We are at a tipping point of creating distrust over trust , disengagement over engagement, complexity over simplicity, wastage over value, loss over saving.
Personally I want the data world where we are all fully engaged and sharing data. High on a mountain with a beautiful view.
In summary
Data will understand a wink, our challenge is keeping users engaged and create usability based on fabulous user experience . This I believe is our industry’s biggest challenge and those who get the balance right between collecting data, engagement and creating value .... will win .... but we need to all work together and make user experience rock. One bad egg, one too greedy ... an industry of disengagement
My closing challenge - Which one of you professionals here at "Digital Shoreditch" is going to make that difference?