The plethora of screens, gadgets, wearables and other technology in our lives is visible and, in some ways, trivial. The reliance on them, though, is a concern.
We’ve made a God-shaped hole and filled it with technology. We rely on statistical prediction far too much. We’ve forgotten the value of individuals.
Timandra Harkness – comedian, author of several big data books, and presenter on various BBC radio shows relating to data science and the world around us – is concerned.
Big data: Difficult point
Primarily interested in the things that have an impact on society more widely, Harkness is honing in on “where technology meets humans at its most difficult”, as we talk about her upcoming appearance at SCI:COM in Dublin in December.
“Big data has immense potential in science and industry and doing things more efficiently,” she said, “but when people use it to understand human beings, I find that more difficult.”
Scientists and researchers can make a mathematical model of what society is, or what people are doing, and predict general patterns. However, according to Harkness, there are “human reasons” as to why this is sometimes wrong.
“Saying why the average individual ‘will do this’ is different to saying an actual individual will do something,” she said. “I found a basic online tool that predicted my death. It said 84 years of age. I put in a posher post code from nearby and I got 89. We all know that’s nonsense.”
Realising these models are statistical, and nothing more, is important to Harkness. However, admitting that they can be useful is key.
Big data: Healthy planning
If, for example, you run a hospital and the data shows you are using ‘x’ amount of Zimmer frames or crutches each year, this can help you plan for future purchasing. “It’s useful for broad brushstroke predictions,” said Harkness.
“But when people start thinking of big data as an oracle that knows the future as magical and high tech – we’re in trouble. We’ve made a God-shaped hole…”
One of Harkness’ many BBC projects, called Singularity, deals with this in far more detail. Her talk at SCI:COM will no doubt touch on this specific area of big data, but her background hints at more of a broad grasp on the general subject.
Harkness wrote a book recently called Big Data: Does Size Matter? – purely designed for a reader whose first question might be ‘what is big data?’.
Her work with numerous BBC radio programmes on this theme came after a stint as a stand-up comedian, joking about parties, being drunk and everything else you expect to hear from a comic. Though stand-up eventually bored her.
Now emotionally and professionally invested in data science, her modern comedy shows – alongside Matt Parker – fall down on mathematical and scientific lines.
Your days are numbered: The maths of death is one such show, with risk, epidemiology, numbers, stats, engineering, brain science and gender regular themes.
Big data: Who keeps up?
Broaching areas like this with an auditorium full of people there to see a talk, or a venue full of people looking for a laugh, may seem a bit obstructive. Though Harkness doesn’t see it that way.
“Science fiction really helps with communicating,” she said, with the likes of Minority Report, Westworld, iRobot, Snowden, AI and other major films and TV shows driving home the reality that data is everywhere, and it must be understood.
“The nice thing about doing live work is you can ask people, ‘Who has heard this before?’. Then you just read the room.
“There’s a huge range of what people know and enjoy. Some are really interested in science and technology, science fiction, nature or whatever. They know quite a lot about what is going on.
“Others don’t follow it closely, or feel confident enough to ask what the hell you’re going on about. But because I’m no scientist, I’ve never been too embarrassed to ask basic questions like this.
“People may not know stuff, but they’re not stupid. They can keep up once they’re told. So I tell,” Harkness said.
Timandra Harkness will speak at the SCI:COM Science Communication conference in Ballsbridge Hotel Dublin on 7 December. More information on www.scicom.ie.