Samsung, Big Data & Dumb Heuristics

Gerd Gigerenzer is a strong advocate of the idea that simple heuristics can make us smart. Yet heuristic might be the root cause of our deep fear of big data, big brother and our data being misused. Imagine you’re a turkey at Christmas. Every day you are approached by a man with a bucket of corn who feeds you. What kind of mental model of what happens when he appears, do you think you will build up? Gigerenzer is the Director of the Max Plank Institute for Human Development and might use this story to demonstrate how high tech firms approach big data to predict the future from past decisions. These organisation make predictions based on the past that are only ever correct by chance and we believe them in order to absolve ourselves of responsibility for when things go wrong. According to many it’s a waste of time, money and talent. Because systems will always go wrong. Because, of course, Christmas always comes.

In psychology, heuristics are simple, efficient rules which people often use to form judgments and make decisions. They are mental shortcuts that usually involve focusing on one aspect of a complex problem and ignoring others. This occurs when you talk with groups about Big Data. People’s gut reactions (heuristics) to things tell them that information about themselves or their loved ones, collected and disseminated, is bound to be used for evil rather than good. But how valid are their concerns about big data and this digital collection of our thoughts? Collecting information on us might fuel the fear of big data. Or as in this brilliant tweet, suggesting Samsung SmartTV instruction manual is taken from George Orwell’s 1984.

The Samsung revelation on Monday 19th February 2015 in the Daily Telegraph,

‘Samsung SmartTV customers warned personal conversations may be recorded. Voice recognition software could transmit ‘personal or sensitive’ information Families are being warned that modern televisions are recording their conversations and could transmit the messages to “third parties”. Many of the latest sets have microphones so viewers can change channel, turn on a DVD or browse the internet by speaking at the screen or remote.But the small print of the privacy policies for these so-called Smart TVs contain warnings that general conversations are also being recorded.Television companies advise users who are concerned to avoid discussing “personal” matters in their livings rooms.The practice, which emerged on internet forums yesterday, led customers on social media to draw comparisons to George Orwell’s 1984. It is unclear whether the information is used for marketing purposes or held on computer that could be hacked by criminals.’

Count on the great Isaac Asimov to have presaged it, much like he did online education, the fate of space exploration and even the rise of Smart TVs. In his legendary Foundation trilogy, Asimov conceives his hero, Hari Seldon, as a masterful mathematician who can predict the future through complex mathematical equations rooted in aggregate measurements about the state of society at any given point in time. Like Seldon, who can’t anticipate what any individual person will do but can foreshadow larger cultural outcomes, big data is the real-life equivalent of Asimov’s idea, which he termed psychohistory, an invaluable tool for big-picture insight into our collective future.
But here comes the irrational fear again. The human brain comprises two distinct parts (if not 3): the old simple brain in the back, which produces impulses and instincts that help us to survive, and the new brain in the front of the head, which we use to control those impulses.The problem is that, sometimes, the two conflict. For example, we might crave fast food even while knowing it’s bad for us. Or we might feel fear when standing on top of the Shard even though we know we’re not going to fall. New research challenged the idea that human beings are rational actors, but provided a theory of information processing to explain how people make estimates or choices.

Heuristics mean we (traditionally) don’t need complex models of the world to make good decisions. These rules work well under most circumstances, but they can lead to systematic deviations from logic, probability or rational choice theory. The resulting errors are called cognitive biases and many different types have been documented. Heuristics usually govern automatic, intuitive judgments but can also be used as deliberate mental strategies when working from limited information.

No matter how much I know about heuristics and concentrate on avoiding the sunk cost fallacy, I still naturally gravitate towards it. The term sunk cost refers to any cost (not just monetary, but also time and effort) that has been paid already and cannot be recovered. So, a payment of time or money that’s gone forever. The reason we can’t ignore the cost, even though it’s already been paid, is that we wired to feel loss far more strongly than gain. Psychologist Daniel Kahneman explains this in his book, Thinking Fast and SlowOrganisms that placed more urgency on avoiding threats than they did on maximising opportunities were more likely to pass on their genes. So, over time, the prospect of losses has become a more powerful motivator on your behaviour than the promise of gains

The sunk cost fallacy plays on this tendency of ours to emphasise loss over gain. A brilliant 1985 (nearly 1984) research study by Hal Arkes and Catherine Blumer is a great example of how it works. They asked subjects to assume they had spent £1000 on a ticket for a ski trip in the Switzerland, but soon after found a better ski trip in France for £500 and bought a ticket for this trip too. They then asked the people in the study to imagine they learned the two trips overlapped and the tickets couldn’t be refunded or resold. Which one do you think they chose, the £1000 good vacation, or the £500 great one?

  • Over half of the people in the study went with the more expensive trip. It may not have promised to be as fun, but the loss seemed greater.

The sunk cost fallacy leads us to miss or ignore the logical facts presented to us, and instead make irrational decisions based on our emotions (see the news and big data fears) without even realising we’re doing so. The fallacy prevents you from realising the best choice is to do whatever promises the better experience in the future, not which negates the feeling of loss in the past.

Could it be that our fear of Big Data emanates from this short cutting old brain structure? What exactly are we really afraid of? If big data can help us overcome these heuristic brain errors, it holds the promise of righting the balance of quality over quantity in our culture of information overabundance, helping us to extract meaning from (digital) matter. Just as the old brain interprets images in art and film as being real (this is why we feel afraid when watching horror movies) perhaps it conflates our hypothetical big data fears with reality.

Our best bet is to try to separate the current facts we have from anything that happened in the past. In a society that tweets more words every hour than all of the surviving ancient Greek texts combined, we certainly could use that.

Samsung can only hope we can separate the two very quickly.

Be Amazing Every Day.