(2019) Chris Lintott, Oxford University Press, hrdbk, £20, xv + 265pp, ISBN 978-0-198-84222-4
It may seem paradoxical to begin a review by talking about another book altogether, but Chris Lintott’s The Crowd and the Cosmos is best approached with reference to James Lovelock’s latest book (with Bryan Appleyard), Novacene: The Coming Age of Hyperintelligence (Allen Lane, 2019). Part of Lovelock’s thesis is that the age some call the Anthropocene, the period in which humans have attained the power to reshape the planet, is already coming to a close, and has not been such a bad thing as often made out, because it has given rise to the new forms of artificial intelligence which will characterise what he tentatively calls the Novacene – a new age which will transform not just the Earth, but, in effect, the entire Universe and where we fit into it. It’s not something to be feared, he argues: the machines will not wipe out life on Earth, because they need the same conditions as we do to survive, and those conditions are dominated by the needs of life and the planet-changing steps it has already taken, and must continue to take, to satisfy those needs, including the need to evolve. There is no self-aware grand plan for ‘Gaia’, but life behaves as if there were one. We and the machines will have to find our place in it, together. When we do, we will not be the dominant partners – indeed, they may prevent us from the worst kinds of folly – but we will have our valued place, because of the things we can do better than they can, especially in concert with them. And although Lovelock doesn’t mention Lintott’s ‘Zooniverse’, this book is about just exactly that.
It takes a while to get there. The first quarter of the book is spent analysing what science is and how it’s done, a slippery subject indeed. His take on it might seem like a scattergun approach, but what he’s doing is to define the Zooniverse is by exclusion – ‘the Zooniverse is any of the above, it’s a new way of doing things’. The particular skill that humans have, and machines are not particularly good at, is discrimination. Get a lot of humans doing that, with modern communications skills and the power of computers, all concentrating on the elements of a very large data set, and amazing results appear.
For Lintott, it all began at a Royal Astronomical Society event, with the launch of a project called Galaxy Zoo. The object was to harness human skills to sift through the huge volume of images collected by the Sloan Digital Sky Survey, and sort the galaxies from the star clusters and the nebulae, then classify the galaxies by types, while taking note of, but not distracted by, other characteristics such as colour. By the time Lintott has explained why this was important, we’re nearly a third of the way through the book; but the key point is that it took only a mention on the BBC News website to evoke a response which swamped first the emails, then the servers supplying the data to the emailers, und so weiter.
It’s characteristic of astronomy that the leg-work, compiled for instance by the various sections of the British Astronomical Association (lunar, solar, variable stars, etc.), is done by amateurs. This is not always understood even by other branches of science: some years ago a serious proposal for an amateur observatory in the Falkland Islands, backed by Patrick Moore among others, was turned down by the Islands’ government because the British Geological Survey told them no worthwhile work could be done by non-professionals. But the Galaxy Zoo was in a different league, getting large numbers of people with no previous knowledge, or at best self-taught, to look at great numbers of images and use their own judgment to list what they contained. Not only did it work, but as the compilations of over 100,000 categorised images built up, it became possible to use them for previously unsuspected lines of research. And then people began going beyond the list of categories they’d been given. It’s been said that the most important words in the history of science, many times over, are, “That’s odd…” And it’s been remarked with concern that things like Jocelyn Bell’s discovery of pulsars might never happen nowadays, because if the print-outs she was studying had been evaluated first by computers, the ‘scruff’ that she noticed would have been dismissed as interference, or unimportant. But the people taking part in the Cosmic Zoo began saying, “That’s odd…”. A spectacular example was Dutch schoolteacher Hanny van Arkel, whose ‘Voorwerp’ proved to be the first of a previously unknown class of gas clouds illuminated by sporadic events in nearby galaxies, Then came Green Pea Galaxies, a new type characterised by oxygen emission from star formation on an unprecedented scale.
The success of Galaxy Zoo led to many more possible applications for ‘citizen science’, and by 2008 Lintott was receiving funding, first from Microsoft, then from the Leverhulme Trust, to develop the Zooniverse – “a platform consisting of many projects, rather than just a single website”. Solar Stormwatch, to detect the first signs of Coronal Mass Ejections and track them across the Solar System, was an early example, followed by Old Weather, compiling meteorological data from old ships’ logs, Spacewarps, looking for gravitational lensing, and then studies of plankton behaviour, animal movements on the Serengeti, penguin behaviour, particle physics, searching for distant supernovae, exoplanet transits, updating maps after tropical storms, “and many, many more”.
Readers well versed in astronomy may find the detailed explanations over-familiar, and read the book as I first did, looking for titbits of information that I didn’t already know. But that isn’t the point: Chris Lintott has created something which has brought huge numbers of people into participation in astronomy, and his main aim is to explain to them just what it is they have accomplished. In that, I believe it will succeed very well.
One remarkable effect of the process has been that data from ‘citizen science’ has become acceptable in scientific literature, needing neither explanation nor apology. Lintott highlights another in his Preface:
“at a time when we tend to talk about the internet and the communications revolution it has precipitated in mostly negative terms, it’s a reassuring reminder that the vast majority of people, both individually and collectively, are good. Even when assembled as that most modern of bugbears – a crowd on the internet – they are capable of remarkable feats of both generosity and skill.”
Looking into the future in his final chapter, he foresees “a universe in which we don’t need to reply on advances in machine learning to get the best out of the wealth of scientific data that we now have access to. It’s one where human intuition and pattern recognition are still needed to get the most from data, even when machines are good at classification themselves… There will be a human element to it for a long time yet.” And if James Lovelock is right, that will be only the beginning of the great partnership to come.
Duncan Lunan
[Up: Non-Fiction Index | Top: Concatenation]