Ever since the beginning of the artificial intelligence (AI) boom in the early 2010s, there’s been a corresponding drought in talented AI developers and researchers. The way to fix this is to educate more of them, and last week, MIT announced a $1 billion initiativeto do exactly that: it will establish a new college of computing to train the next generation of machine learning mavens. 

Importantly, the college isn’t just about training AI skills. Instead, it will focus on what MIT president L. Rafael Reif calls “the bilinguals of the future.” By that, he means students in fields like biology, chemistry, physics, politics, history, and linguistics who also know how to apply machine learning to these disciplines.

Two-thirds of the planned $1 billion commitment has been raised so far, with $350 million coming from Stephen A. Schwarzman, CEO of the private equity firm Blackstone. The new college will be named after Schwarzman, and it will include 50 new faculty positions, half of which will focus solely on computer science, while the rest are jointly appointed by the college and other MIT departments. The college is scheduled to open in September 2019, and its new building is planned to be completed in 2022. 

MIT is also angling the college as an ethically minded enterprise; one of its stated aims is to research “ethical considerations relevant to computing and AI.” It’s a frequent criticism of contemporary AI efforts that researchers sometimes ignore the history and lessons of the fields they are trying to “disrupt.” More collaboration should help. 
​
“As computing reshapes our world, MIT intends to help make sure it does so for the good of all,” said Reif in a press statement. “The MIT Schwarzman College of Computing will constitute both a global center for computing research and education and an intellectual foundry for powerful new AI tools. Just as important, the College will equip students and researchers in any discipline to use computing and AI to advance their disciplines and vice-versa, as well as to think critically about the human impact of their work.”

A new method of analyzing images from CT scans can predict which patients are at risk of a heart attack years before it occurs, researchers say.

The technology, developed by teams at Oxford University and institutions in Germany and the United States, uses algorithms to examine the fat surrounding coronary arteries as it shows up on computed tomography (CT) heart scans.

The new heart attack warning system, called the Fat Attenuation Index (FAI), was tested in a large study published by The Lancet medical journal.

The study monitored the progress of 3,900 heart patients from Germany and the US for ten years after they had undergone a coronary CT scan or angiogram.

FAI predicted fatal heart attacks many years before they happened.

Fat gets altered when an artery becomes inflamed, serving as an early warning system for what one of the researchers believes could be up 30% of heart attacks.

“If you can identify inflammation in the arteries of the heart then you can say which arteries will cause heart attacks,” Oxford Professor of Cardiovascular Medicine, Charalambos Antoniades, said.

“With the new technology, we can achieve this by analyzing simple CT scans.”

Most heart attacks are caused by a build-up of plaque – a fatty deposit – inside the artery, which interrupts the flow of blood.

Currently, CT scans tell a doctor when an artery has already become narrowed by plaque.

With the new technology, for which the researchers hope to gain regulatory approval on both sides of the Atlantic within a year, doctors will be able to say which arteries are at risk of narrowing. 

“(We) can say your arteries are inflamed, and a narrowing will be developed five years down the line. So maybe you can start preventive measures to avoid this formation of the plaques,” Antoniades said.

Heart disease and stroke are the two most significant causes of death worldwide.

“Although we have not estimated the exact number of heart attacks that we can prevent, we could potentially identify at least 20% or 30% of the people before they have (one),” Antoniades said.
​
An Oxford University spin-off company is now developing a service to analyze CT scans from across the globe in around 24 hours.

Many of us spent our college years sitting in countless lectures accompanied by uninspiring PowerPoint presentations. The notes we took were equally uninspiring and weren’t remembered.

It’s a problem that doesn’t just affect students. Have you ever caught yourself re-reading the same article twice? To make text easier to remember, Melbourne-based designers partnered with behavioral scientists at RMIT University and have designed a font that makes reading more difficult.

Ironically named “Sans Forgetica,” the font is more difficult to read than most typefaces – and that’s by design. The ‘desirable difficulty’ you experience when reading information formatted in Sans Forgetica prompts your brain to engage in deeper processing.

To make your brain work harder to retain information, the font is legible yet broken and disconnected. Since chunks of words are removed, the brain is required to put more effort into reading, which only takes a fraction of a second – but this font seems to be working.

A recent study involving 400 university students found a small increase in memory retention when reading text in Sans Forgetica compared to Arial. The participants reading Sans Forgetica were found to remember 57% of the text, and only 50% when reading in Arial.

Although the font has promising statistics, it has limitations according to lecturer and Sans Forgetica co-creator, Stephen Banham. “You wouldn’t want novels printed in it. It would probably induce a headache,” Banham said.
​
It’s unlikely you’ll want to use it as part of your every-day reading, Sans Forgetica could be a handy type-tool for stressed-out students. You can download this free font here.

Google frequently touts that the “next billion users” will come from developing nations with different focuses and needs. To that end, the company has developed several optimized services, with one being a consistent and straightforward addressing system that works globally.

Google created “Plus Codes”for addresses that are not easily located through conventional descriptors. First introduced in February 2018, Plus Codes allow users to locate almost any spot on the planet. What many people in traditional civilizations fail to realize is that, according to a World Bank estimate, half of the world’s urban population lives on unnamed streets.

Notably, this open source solution composed of 10 characters works globally and can be incorporated by other products and platforms for free, with a developer page available here.

It works offline and in print when overlaid as a grid on existing maps. Places that are close together share similar plus codes, while the system is identifiable by the “+” symbol in every address.

This system is based on dividing the geographical surface of the Earth into tiny ‘tiled areas,’ attributing a unique code to each of them. This code comprises a ‘6-character + City’ format that can be generated, shared and searched by anyone – all that’s needed is Google Maps on a smartphone.

                   •   The first four characters are the area code, describing a region
                        of about 100 x 100 kilometers.

                   •  The last six characters are the local code, describing the neighborhood
                       and the building, an area of approximately 14 x 14 meters – about the
                       size of one half of a basketball court.

The area code is not needed when navigating within a town, while another optional character can be appended to provide additional accuracy down to a 3 x 3-meter region. Plus codes for any location can be found with this tool.