The human brain really is an amazing piece of engineering. A small organ the size of two fists that reasons, analyses, manages emotions, regulates blood flow and controls our temperature. It instantly interprets danger from safety, fake smiles from real ones, surprise from anger and allows us to select our weekly groceries from a choice of 40,000 in only 30 minutes. And it does all this using around a third of the power of a 60-watt light bulb.
Because it’s such an amazing piece of engineering, it’s often compared with other great feats of engineering and technology, like a water clock, a telephone switchboard and, more recently, a computer. Today’s analogy spawns quotes like “the brain can process at the rate of 168,000 pentium computers” or “the brain has about 100 million megabytes of memory”. On their own, these quotes are fine, but if we carry the analogy too far we start thinking the brain actually works like a computer.
While brains and computers both have memory and processing power, the brain is designed to work differently. This is why humans are better than computers at some stuff, and worse at others. We don’t file information anywhere near as well for a start (something we can all probably relate to when we lose our car keys, can’t recall a friend’s name or inevitably forget things at the supermarket without a list). We also can’t retrieve our memories as well as a computer. Questions like “are there more four letter words with R as the first letter or the third letter?” are a piece of cake for a machine but not so much for us. We want to say there are more words with R as the first letter because that’s how the brain files this information, but the answer is the third letter (and by quite some way).
Computers are much better at detailed analysis and can retrieve information easily in a variety of forms. That’s why they now beat us at chess and why Google beats Encyclopedia Britannica. But put the world’s best robot on a tennis court and even average players would win in a canter.
We’re not particularly good at estimating odds. We use an ‘ease of imagining’ metric as a ballpark of the likelihood of an occurrence, which means we over-estimate dog attacks and under-estimate asthma fatalities. That’s because dog attacks make the news (making them easy to recall) and asthma attacks don’t.
Unlike a computer, the brain was not optimised for analysis, it was optimised for efficiency. From an evolutionary point of view, a reasonable decision made quickly was considerably more beneficial to our survival than a slower, more accurate one.
This is why we have a number of analytical glitches (known more commonly as cognitive biases). Of the 11 megabits of data that enter the human brain each second, we consciously process only around 40 bits which means we subconsciously process 99.9993 percent of the information around us. It’s an extremely efficient way to operate and most of the time we don’t notice any negative effects. But upon closer inspection of these data compressions, short cuts and approximations the brain is making, we can see a few errors creeping in.
For example, we’re not particularly good at estimating odds. We use an ‘ease of imagining’ metric as a ballpark of the likelihood of an occurrence, which means we over-estimate dog attacks and under-estimate asthma fatalities. That’s because dog attacks make the news (making them easy to recall) and asthma attacks don’t.