Share #TechAways with your friends and colleagues. Ideas? Suggestions? Comments? Send them our way!


Tech Team

Tech is taking on the EU’s darling industry[Fast Company]

What if tech could cut the Common Agricultural Policy from a third of the EU budget to nothing? Matt Barnard, co-founder of the indoor vertical farming start-up Plenty believes his technology can grow up to 350 times more produce in the same amount of space as conventional farming with only 1% of the water. This means more efficient production but also tastier products, shorter supply chains and less pollution from farming. And we’re not just talking salad here but strawberries, maybe tomatoes, cucumbers and more. Will the European agro-industry adapt or will policymakers protect their darling industry?

Bias, recorded. [Harvard Business Review]

We are surrounded by implicit bias, but rarely get to see it quantified. A new study out of Sweden is changing that. A team of researchers recorded government venture capital decision-making meetings and transcribed the discussions over a two-year period. Not surprisingly, they found that male and female applicants for funding were described in radically different terms. See a graphic here detailing the difference. That biased language resulted in women entrepreneurs only being awarded 25% of the applied-for amount, whereas men received 52% of what they asked for. Luckily, the authorities have taken action – hopefully a data-driven trend that we will see continue.

The big bad lack of information in health data sharing. [Sky News]

Cyber-attacks are not the only concern for the UK’s National Health Service (NHS) as data protection is again on everyone’s minds. A Sky News investigation claims that more than 1.6 million identifiable patient records were inappropriately shared with Google’s DeepMind for the testing of a new eHealth app. Experts say there were “fundamental errors” in this data sharing project, and DeepMind admits more could have been done to inform the public about how their data is used. While the UK’s data watchdog looks into the legality of the transfer, it may be time for the NHS, and public institutions across Europe, to reconsider how they handle and protect user data.

Too much of a good thing? [Computer Weekly]

Proprietary software is great, and it has without a doubt changed the world over recent decades. That said, many in Europe are worried that an overreliance on one provider might increase costs, stifle innovation, and raise security risks. As public IT systems including tax, health, or policy records, continue to digitise and move to the cloud, some authorities are pondering a move to open source. Check out this long read for the latest on the ground examples of the debate from the UK to Italy.

What worries people about machine learning? [The Guardian]

The Royal Society published a report on perceptions of machine learning and AI. Despite some concerns about possible physical harm from embodied systems such as autonomous vehicles or social care robots in the home, the general opinion is that machine learning could enhance lives in ways from advantages in healthcare to efficiency in public services. People were even optimistic about the effect on their jobs, citing ideas for how machine learning could enhance rather than replace existing roles. The European Commission recently released a similar studyof EU citizens’ views on digitisation and automation. Shockingly, or not, 75% or respondents thought the impact would be positive. It seems AI and robots haven’t made an enemy of us yet.

A Silicon-built future. [The New York Times]

R&D spending is essential for driving innovation. But US government funding for scientific research and technology is falling and the news coming out of the White House suggests this trend may continue. Left to fill the gap are technology giants. As the big five, Google, Amazon, Apple, Facebook and Microsoft continue to invest heavily in AI and other technologies that will radically change our societies and how they govern, should it cause alarm that public research isn’t playing a larger role?