by Patrick Worms
Every single day, millions of land users around the world take decisions that will eventually threaten them, their children, their neighbours and millions of strangers around the world with poverty, destitution and death by starvation, extreme weather or disease.
The collective impact of these decisions drive deforestation, land degradation, desertification and biodiversity loss. And yet hardly any of the people taking these decisions know exactly the damage they are doing. Nor do they know of any other way of achieving their objectives.
They are merely people struggling to make a living, applying very limited knowledge to their particular circumstances.
In that context, ensuring that the technology packages recommended to individual farmers and other land users are as near to optimal as humanly possible is absolutely crucial. And so is figuring out ways of getting that knowledge to the hundreds of millions of land users who need it, irrespective of their economic status, linguistic abilities or literacy levels.
The ongoing investigations of which particular agroecological technology packages are optimal in specific geo-socio-biophysical environments are part of that work. Co-learning, as recently described on this blog (“One small change of words – a giant leap in effectiveness”), is essential to that effort.
But getting that information to users is perhaps an even bigger challenge. That is why some envision the development of a sophisticated electronic expert platform capable of offering any land user tailored access to everything known that could help them: optimal options considering their soils, climate, the sizes of their farms, the cash and labour they can afford to invest, the local markets that can be accessed and even the policy issues that will influence their choices.
Yes, literally “everything”: a distillation of the millions of pieces of data, of social and scientific evidence that have been gleaned about various social and agroecological systems over the decades by hundreds of institutions and tens of thousands of researchers, development practitioners and journalists publishing in many languages around the world.
Imagine a tool built on huge databases, constructed with sophisticated algorithms designed to make finding the right information elegant and painless for users. A tool built to be as easy to use as possible, and to be accessible to practitioners in a multiplicity of languages through a multiplicity of devices. A tool that can access, store, analyse, translate and deliver information lifted from millions of papers published in dozens of languages. A tool that is clever enough to be accessed even by the semi-literate using the simplest mobile phones.
Such a tool would be clever enough, and have access to enough information, to suggest literally thousands of different agroecological systems to land users depending on their unique local circumstances and needs. It would be co-learning by machine, on a vast scale.
Utopia? Today, yes. But it is not an impossible job. The open data movement is successfully forcing hundreds of institutions to make their information available to anyone who wants to use it. Google and Microsoft and others constantly improve their extraordinarily effective search algorithms. News services already publish machine-generated articles constructed from sophisticated automatic web searches. The data gathered from the daily habits of the billion plus cellphone users is already revolutionising fields from crowd or traffic management to epidemic forecasting. Thousands of activists around the world are developing a cornucopia of smartphone tools for public benefit. And Google Translate already allows me to discharge my duties as a trustee of a Hungarian NGO, despite my total ignorance of Europe’s most isolated language.
Today, most of what is know about agroecology is in effect inaccessible. It could be carved in Egyptian hieroglyphics on porphyry slabs buried under 20 feet of sands, for all that 99% of the potential users of that information know. Even those, like ICRAF scientists, who are in the thick of it have little access to data and papers published more than a few years ago, or that appear in papers that neither they nor their professional contacts read, or that was published in languages they do not master.
In effect, much of the huge effort that tens of thousands of people made over decades to document and understand agroecological systems is wasted. Lifting that veil of ignorance is the challenge such an expert platform would tackle.
What is missing is not the technological capability to build that platform, but something else: an awareness among those that drive the development of these technological capabilities that the issue needs tackling urgently.
As a consequence, this expert platform is not even being conceptualised. And tomorrow, millions of people will again, through blissful ignorance, sap our collective ability to survive the crises the future will bring.
Many of the titans of the tech world are deeply worried about the global future. They use their considerable wealth to finance efforts to tackle some of the world’s least tractable problems, often through technology-centric initiatives. Imagine if they became as fired up about the scandal of hidden agricultural data as they are about other issues!
That will never happen if they don’t know the problem even exists, or why it matters.
They don’t know the problem exists because we in the agricultural research in development community have lived with the fact that most information is “dark” all our professional lives. We have become habituated to this scandal. We treat it like cancelled flights: an irritation perhaps, but one not worth wasting time talking about seriously. So we rarely if ever speak about it or mention it in our statements, papers, visions or policy briefs. And by “we”, I mean the entire development and research community (scattered efforts are probably looking at these issues, a few papers may well have been published. But they, too, are mostly “dark” to most of us: we are agriculturalists, after all, not data technologists).
We don’t get outraged at this situation, and nor do we get our partners outraged. And yet it is an outrage. That information has the potential to transform global agriculture, and thus save countless lives today and in the future.
We can’t build that platform. The entire CGIAR working with the big 5 NGOs and a consortium of big donors could not build that platform. But the likes of Google or Oracle can. They have the technologies, the brains, the money and the relationships. And in many cases, their owners have the vision to tackle the world’s most intractable issues.
So let us make them aware of the huge damage that dark data is doing to the fabric of our world. Let us trumpet to the skies that we need their help to reveal that data to everyone who might need it.
Let us all use every opportunity to talk about it. Let us condemn the scandal of dark data in our papers, our conference conclusions, our statements, articles and presentations. Let us call for open, accessible data whenever we can. Let us link across institutions to push the case. Together, we can build a head of steam that will, eventually, reach the ears of those who can tackle this challenge.