Innovation & Technology
Weekly Roundup

This is the online version of the latest UNU-MERIT I&T Weekly which is sent out by email every Friday. If you wish to subscribe to this free service, please submit your email address in the box to the right.

This week's headlines:



Google Translate AI invents its own language to translate with
November 30, 2016

Google Translate is getting brainier. The online translation tool recently started using a neural network to translate between some of its most popular languages - and the system is now so clever it can do this for language pairs on which it has not been explicitly trained. To do this, it seems to have created its own artificial language.

Traditional machine-translation systems break sentences into words and phrases, and translate each individually. In September, Google Translate unveiled a new system that uses a neural network to work on entire sentences at once, giving it more context to figure out the best translation. This system is now in action for eight of the most common language pairs on which Google Translate works.

Although neural machine-translation systems are fast becoming popular, most only work on a single pair of languages, so different systems are needed to translate between others. With a little tinkering, however, Google has extended its system so that it can handle multiple pairs - and it can translate between two languages when it hasn't been directly trained to do so.

For example, if the neural network has been taught to translate between English and Japanese, and English and Korean, it can also translate between Japanese and Korean without first going through English. This capability may enable Google to quickly scale the system to translate between a large number of languages.

Google's researchers think their system achieves this breakthrough by finding a common ground whereby sentences with the same meaning are represented in similar ways regardless of language - which they say is an example of an 'interlingua'. In a sense, that means it has created a new common language, albeit one that's specific to the task of translation and not readable or usable for humans.

Full story: New Scientist Back to top


One simple algorithm could explain human intelligence
December 01, 2016

It's a mind-bending idea: that all the complex thoughts running through our heads are the product of a set of definable sums. But scientists have identified clear patterns in the brains of mice and hamsters, and if a similar phenomenon could be found in human brains, it could form the basis of such an algorithm for intelligence.

Last year, Joe Tsien from Augusta University in Georgia published his 'Theory of Connectivity', in which he put forward the idea that groups of neurons, called neural cliques, come together in pre-wired ways to process thought and knowledge. Essentially, it's a framework for arranging the brain's billions of neurons. Tsien says these cliques then assemble to form functional connectivity motifs (FCMs), which could represent all the potential variations in human thought. The harder we need to think the more cliques are required to make up that FCM.

For his latest study, Tsien put his Theory of Connectivity and FCMs to the test, using electrodes implanted at specific points in the brains of mice and hamsters to monitor neuron activity. His team was able to predict the neural cliques that formed in response to certain scenarios, such as the arrival of food or the presence of a threat. Depending on the scenario, the animals' neurons arranged themselves in very predictable groups.

At the centre of Tsien's hypothesis is the formula n = 2^i - 1, where 'n' is the number of connected neural cliques, '2' indicates whether the neurons are receiving an input or not, 'i' is the information being received, and '-1' is accounting for multiple possibilities. Tsien says this formula is enough to predict FCM grouping.

Full story: Science Alert / Frontiers in Systems Neuroscience Back to top


Bill Gates is backing a waterless toilet - here's how it works
December 02, 2016

Four years after the Bill and Melinda Gates Foundation awarded USD 710,000 for the development of a waterless toilet, the technology has received a second financial boost. Researchers at Cranfield University will use the money to continue developing the lab prototype of their Nano Membrane Toilet and begin field trials in Africa.

After a person has done their business and closed the lid, the rotating toilet bowl turns 270 degrees to deposit the waste in a vat underneath. A scraper tool then wipes off any residual waste from the bowl. The solid waste stays on the bottom while the liquid rises to the top.

Extremely thin nanofibres are arranged in bundles inside the chamber. They help move the water vapour that exists as part of the liquid waste into a vertical tube in the rear of the toilet. Next, water passes through specially designed bundles that help condense the vapour into actual water, which flows down through the tube and settles in a tank.

As for the solid waste that's left behind, a battery-powered mechanism lifts the remaining matter out of the toilet and into a separate holding chamber. There it's coated in a scent-suppressing wax and left to dry out. Every week, a local technician visits the community to remove the solid waste and water, and replace the toilet's batteries if needed.

Residents can then use the water for tending to their plants, cleaning their homes, cooking, and bathing. The solid waste ends up at a thermo-processing plant to be turned into energy for the community. One toilet can accommodate up to 10 people for USD 0.05 per day, per user - in line with the Bill and Melinda Gates Foundation's original criteria for the prize. Field testing will begin later this year.

Full story: Science Alert Back to top


Smartphones may be the new tool in fighting mosquito-borne diseases
November 30, 2016

Smartphones can tell one type of mosquito from another by its hum, which may be useful in fighting mosquito-borne diseases, according to new research from Stanford University in the US.

Calling their project 'Shazam for Mosquitoes', after the phone app that identifies music, students from the university's Bio-X institute showed that common phones could record mosquito wing beats accurately enough to distinguish, for example, Culex mosquitoes, which spread West Nile virus, from Aedes mosquitoes, which spread Zika. Even older flip phones are sensitive enough to do the job.

The students envision a crowdsourcing initiative in which phone users around the world send in sound samples of mosquitoes landing on them, which could be sorted by the embedded GPS and time co-ordinates to build a worldwide mosquito distribution map. It would be far less cumbersome than the current technique: trapping insects for hand sorting.

Mosquitoes use their wing-beat hums to find one another for mating. The sounds are distinct, and even big and small members of one species make similar hums. Less than half a second of flight is needed to capture a mosquito's acoustic signature, and the technique works even against background noise like sirens or conversation, according to its developers.

Full story: Sydney Morning Herald Back to top


EU unveils plans to boost 'clean energy' use
November 30, 2016

The EU on Wednesday unveiled 'clean energy' plans to boost renewables, cut waste and reduce subsidies for coal power in a bid to meet commitments under the Paris climate deal.

Binding energy efficiency targets would also be raised by 30% by 2030 under the sweeping package of measures from the European Commission, the European Union's executive arm.

Under the Paris climate deal struck almost a year ago, the EU plans to cut greenhouse gas emissions by 40% by 2030, compared against 1990 levels and make renewable energy account for 27% of energy use.

In 2030, the EU aims to have half of the bloc's electricity come from renewables such as wind and solar power. By 2050, it hopes electricity will be carbon free.

The package calls for making renewable energy increasingly market-based under a regulatory framework. The plan also aims to promote more jobs in a sector that already employs more than one million people, while reversing an economic trend - investment in renewables has dropped by half since 2011.

Full story: Yahoo! / AFP Back to top


MIT researchers discover another strange property of water
December 01, 2016

Despite its abundance, water remains a mysterious substance with a host of strange properties. A new discovery from researchers at MIT just added to the anomalies - even at high temperatures, water freezes solid when placed in tiny tubes.

The discovery came as a surprise to the researchers, who had been trying to send electric currents through water in nanotubes. At temperatures of at least 105 degrees Celsius - well beyond water's usual boiling point - the water in their nanotubes froze solid. The degree and direction of the phase transition was way more than they expected.

The discovery may find practical applications beyond adding to our understanding of water. Due to its ability to remain in a stable state through drastic changes in temperature, these water-filled nanotubes may be used as 'ice wires', which could prove to be efficient highways for transporting protons. Water is 10 times better at conducting protons than standard conductive materials.

But there's another mystery here which hasn't been solved - how water enters the nanotubes in the first place. Nanotubes are just a few water molecules thick and are thought to be water-repellant. Strangely, filling them with water proved easy.

Full story: Yahoo! / Nature Nanotechnology Back to top


Radioactive diamonds are turned into long lasting batteries
November 30, 2016

A diamond lasts forever. So why not transform that diamond into a super long-lasting battery? That is the idea behind a new technique developed by researchers from the University of Bristol that transforms nuclear waste into diamond form. These radioactive diamonds are capable of generating their own electrical current, and since the half-life of their radioactive material lasts for thousands of years, so does the power source.

For many years, nuclear reactors in the UK were encased in graphite blocks to help sustain their nuclear reactions. Graphite, like a diamond, is made of carbon, but because of prolonged exposure to radioactivity, the graphite blocks used at nuclear sites also became radioactive. Specifically, their carbon transformed into an unstable isotope, carbon-14. Because carbon-14 has a half-life of 5,730 years, all of that graphite must be safely stored for a very long time.

Now researchers transformed the carbon-14 into diamond form and then encased it with an outer layer of non-radioactive diamond. Because diamond is the hardest substance known, this outer layer keeps the radioactivity from escaping and causing harm. Meanwhile, that inner radioactivity produces a constant charge until it decays, which is not for thousands of years.

These diamond batteries won't produce a large charge. One gram of it would deliver around 15 joules per day, less than an alkaline AA battery. The perk, though, is in how long lasting they are. The batteries could have value for powering pacemakers, satellites and devices and sensors during long-distance space flights.

Full story: MMN Back to top