Many important discoveries and innovations, such as GPS, microchips and supercomputers, are the result of investments by governments and the private sector. Economist Mariana Mazzucato recalled last week, in a lecture about her book Mission EconomyOnce again, the state investment in the Apollo 11 moon mission allowed many other groundbreaking innovations to sprout: from dried food to fireproof materials.
Half a century later, the CEOs of major tech companies are racing into space. That is exemplary of the growing distance between their possibilities and those of most of us, simple souls on earth. Also when it comes to the capacity to innovate, these top executives have an almost unassailable position. That raises concerns about the opportunities for start-ups; the question is whether the dominant tech companies will ever be overtaken. Public interest is at stake in the development of artificial intelligence; if we are not careful, tech companies will become the only drivers and knowledge carriers.
Huge mountains of data are needed to train the machine models underlying artificial intelligence. Big Tech in particular has access to that data and can pay for the computers with the necessary computing capacity. A single lab researcher, the average European university, even a country like the Netherlands can hardly offer a competitive alternative.
The computer science students I teach at Stanford therefore prefer a job at one of the big tech companies – purely because they have access to the aforementioned data and computing capacity there, and nowhere else. Companies with the largest datasets in the world, such as Google, Facebook, Microsoft and Amazon, are speeding ahead in the field of artificial intelligence. Academic research and innovation that serve the public interest are lagging behind.
Tech companies keep their data sets and algorithms secret and protect them by relying on intellectual property. The gap is growing between what a handful of companies amassed in knowledge and capacity and what is happening in this area in the public domain. Even those who want to academically research or check the latest technologies can’t come close to the data, computing power and expertise that the private sector has collected.
To restore the balance and provide scientists with better access to data, an alternative is now being developed at the initiative of Stanford: a national research cloud. This should lead to academics and civil society organizations gaining access to massive, public datasets. By opening up government data and computing capacity within a shared computer infrastructure, top American universities are working on the democratization of research. The US Congress supports the initiative.
Ursula von der Leyen, President of the European Commission, was right when she said in her State of the Union address on Wednesday that digitization can make or break the EU. The same applies to the Netherlands, where the enormous task of managing digitization in the right direction has been rather overlooked.
At the end of this month, directors and senior officials of the Trade and Technology Council, a consultative body of the EU and US, will discuss important topics such as technology standards, control of exports of dangerous technology, and human rights in the context of digitalisation. Why not immediately discuss the design of a joint transatlantic public research cloud?
Marietje Schaake writes a column on technology, policy and economics here every other week.
Can the data gap with Big Tech still be closed?
Source link Can the data gap with Big Tech still be closed?