As someone who believes that truth can conquer all I have been thoroughly bemused and frustrated by a lot of what has been on show in the UK and the US over the last few months. It seems that huge numbers of people have decided that as long as they like what they hear they don’t really care if it is true or not.
According to fact checkers of the 331 ‘factual statements’ that businessman Donald Trump made during the US Presidential election campaign 231 were false (http://www.politifact.com/truth-o-meter/lists/people/comparing-hillary-clinton-donald-trump-truth-o-met/).
Last week Trump claimed that the US could see 650,000,000 people moving to the US in a week – trebling the population of the country. Anyone with half a brain must know that can’t be anywhere close to be true (as carefully demonstrated by our friends at Wired (https://www.wired.com/2016/11/take-move-650-million-people-us-one-week/).
This led me to go on a bit of a search for people who can restore my faith in the power of in data and science and two weeks ago I found one of these people in the guise of Professor Yuval Harari (http://www.ynharari.com). He is one of those people who makes you feel inordinately stupid, but also relieved that professors and authors like Harari exist. A regular speaker on TED Harari is worth listening to.
The reason I decided to open this month’s article with a focus on the good professor and not the president-elect however is less to do with Harari’s ability to reaffirm my faith in humanity and more to do with the fact that the professor has some very interesting things to say about the convergence of man and machine, some of the ethics around technological advancement, and how that is going to impact the future development of mankind.
Over the last few years I have been struck by how quick we are to embrace science and technology, often with no real understanding of how it might impact us. The most famous example of this was the splitting of the atom but even at a far lower level, our desire for new and different and more and faster can often lead us into places that we don’t want (or need) to be.
The issue of unintended consequences is an interesting one. When you look at how social media and the digital environment has changed our lives there is plenty to rue as well as to celebrate. Anyone that has children will have concerns about how social media exacerbates pressure and creates unrealistic expectations.
Parenting experts are constantly telling us to put down our smartphones and engage with our children. The fact that so much data is available means it is very difficult for any of us to hide and we unwittingly share so much about ourselves so often.
There is no question that there is both Yin and Yang to the advancement of tech. Personally I am someone who is all for companies being able to understand me and therefore give me what I want/ need. Consumer analytics has now reached a point when companies not only know what I want and when I want it but are now developing tools that can tell them why (which quite frankly means they are one step ahead of me most of the time!).
The latest advancement, which Harari actually touched on in his talk, is the development of machine learning and the advancement of the algorithm. Last year companies that I spoke to were obsessed with Data Scientists but, as we move towards 2017, it has become all about algorithms. Algorithms are what Nicholas Antoniou, CEO of Business Mathematics calls ‘numerical recipes’. Used in the right way they bring together a range of ingredients (data) to automate decision-making. It is not a new thing – Antoniou created his first algorithm back in 1975 – but the arrival of companies like ride hailing service Uber, whose business is essentially built on algorithms, means algorithms has become an obsession for businesses of all shapes and sizes.
Algorithms are the latest example of technology innovation that brings with it the issue of unintended consequences. The example that is given most often is the example of the self-driving car. You can create a set of algorithms that will enable the car to make its own decisions however the decision that a machine will make will not always be the same as the one a human would make. The machine will be built to protect the driver but does that mean that given the choice between hitting a child and swerving and putting the driver’s life in danger, it will choose to hit the child?
Not everything is life or death of course. Sometimes it is simply about whether using algorithms to make a decision is the right thing. As Antoniou puts it: “CIOs today are jumping right into algorithmic decision-making but there is a danger that not enough is being done up front to ensure that it delivers something of value. Just because you can do something doesn’t mean you should. You need to decide whether it is relevant (will it deliver change that moves the needle of the business), whether it has been deployed with rigour (do you have enough data?
Have you thought enough about the consequences? Are the predictions made valid under all the relevant situations?) And is it resonant – does the result make a compelling case for what you need to deliver?”
As algorithms become more and more advanced it is easy to see why many people believe they are the key to the next great leap in technological advancement. Advanced algorithms and machine learning are driving the cutting edge of AI – the creation of machines that can think like humans. The most advanced example of this is Google’s AlphaGo – a machine that goes beyond the calculation of possible permutations to one that learns from its mistakes (https://www.wired.com/2016/03/googles-ai-wins-fifth-final-game-go-genius-lee-sedol/).
There is no doubt that complex algorithms are becoming responsible for more and more of the decisions that are made in the modern world. From enabling grocers to map out the optimum delivery routes to determining how long you wait for a taxi. The question is however how much responsibility is too much and are we really prepared for where this technology can take us?