Summary for those who don’t want to read the whole thing: I’m not sure.

Appropriately enough, given the two-part nature of the creature at which I’ll be looking, this will be the first of two blog posts.

A couple of weeks ago I went to the extremely stimulating and interesting launch day for the BBC’s Data Science Research Partnership. Seeing all the committed and intelligent experts in one place was pretty humbling for a data-science-dilettante-with-humanities-background like wot I is.

I particularly liked Cath Sleeman’s work for Nesta on creative networks and gender balance in film (actors and crew) over 100 years in the BFI archive.

Cath’s was one of a series of nine three-minute ‘lightning talks’. Two miracles here: everyone ACTUALLY STUCK TO THE TIME. Credit obviously to the moderator, but also to all the presenters. It was zippy. Second, in every presentation I wrote down something that I found either interesting or that I didn’t know before. All in three minutes.

So yes, nine excellent, concise, and stimulating presentations.

One in particular caught my attention because it used an analogy I myself have used before, but which I’ve come to feel a bit nervous about – that of the centaur.

Eh? Well, very roughly and with many variations and nuances, it’s the idea that although automation – increasingly that automation facilitated by AI and machine learning – can do many tasks previously carried out by people much better than people can, automation plus people providing cognitive skills at the right point will outdo pure automation by itself. That’s your centaur – part horsepower part personpower. #winning

The common example is usually chess – we’ll come on to why in a bit – with computers beating grandmasters, but a grandmaster-plus-computer pairing beating computers.

The ‘new æsthetic’ artist James Bridle (technology as conceptual artistic space) wrote an article What is Wrong with Big Data, which set out the problem, partly by looking at its application to the ‘aw cuet’ world of Big Pharma:

“Eroom’s law” – Moore’s law backwards – was recently formulated to describe a problem in pharmacology. Drug discovery has been getting more expensive. Since the 1950s the number of drugs approved for use in human patients per billion US dollars spent on research and development has halved every nine years. This problem has long perplexed researchers. According to the principles of technological growth, the trend should be in the opposite direction. 

…

In the last 20 years drug discovery has experienced a major strategic shift, away from small teams of researchers intensively focused on small groups of molecules, and towards wide-spectrum, automated search for potential reactions within huge libraries of compounds. This process – known as high-throughput screening or HTS – is the industrialisation of drug discovery. 

…

Drug discovery is faltering, suggest the paper’s authors, because the pharmaceutical industry believes that brute-force exploitation of data is superior to “the messy empiricism of older approaches”

To combat this, many drug companies are combining HTS with human-led exploration – hi, Chemtech Centaur. He then goes on to explore the same approach with regard to chess. It’s a good article, read it. He’s a good artist (drones, Olympic mascots and airspace, how to trap self-driving cars) too.

In this model, humans are seen to add value to the process of automation. They perform a role of bringing expertise, experience, imagination and insight to prioritise and pick from the heavy processing of large amounts of data and algorithmic pattern-finding. In this model, Kahneman’s System 2 thinking brings value to big data crunching. In the knowledge economy especially, humans only bring value when placed at the right point in the automation chain.

In the next blog post, I’ll look at some difficulties and counter-examples of this point of view and come to a conclusion on whether centaurs are real or not.

Tom Wootton, Product Manager Access Services