If there is one subset of machine learning that spurs the most excitement, that seems most like the intelligence in artificial intelligence, it's deep learning. Deep learning frameworks--aka deep neural networks--power complex pattern-recognition systems that provide everything from automated language translation to image identification. Deep learning holds enormous promise for analyzing unstructured data. There are just three problems: It's hard to do, it requires large amounts of data, and it uses lots of processing power. Naturally, great minds are at work to overcome these challenges.

We've entered an era when computers can understand speech, computers can synthesize speech, computers can develop music, author encryption algorithms, create novel art, respond to customer support questions, and even generate new summaries and reports from data. Increasingly, humans will struggle to distinguish between computer-generated and human generated. Consequently, here's an opportunity for startups to lead, not just technologically, but more broadly. DT is a creative digital agency business based in Australia that developed a prototype that distinguishes between human speech and synthetic speech. It's one of undoubtedly many technologies which will use one machine learning model to detect another machine learning model.

In March this year, PWC released a report saying that 10 million UK jobs are at risk of being replaced by AI within 15 years. Prior to this in November 2016, the Bank of England said that AI posed a risk to almost half those employed in the UK and that a "third machine age" would hollow out the labour market, widening the gap between rich and poor. In a recent report the McKinsey Global Institute said that by 2025 AI will do the jobs of 140m knowledgeable and skilled workers globally. Let's not forget the truck drivers. In his speech at Harvard University recently, Mark Zuckerberg said that his generation will have to deal with "tens of millions" of jobs replaced by self-driving cars and trucks. In late May 2017, the International Transport Forum published a report warning that of the 6.4 million professional drivers that are projected to be needed in the US and Europe by 2030, that 4.4 million will be replaced by self-driving trucks.

Popular approaches to differential privacy, such as the Laplace and exponential mechanisms, calibrate randomised smoothing through global sensitivity of the target non-private function. Bounding such sensitivity is often a prohibitively complex analytic calculation. As an alternative, we propose a straightforward sampler for estimating sensitivity of non-private mechanisms. Since our sensitivity estimates hold with high probability, any mechanism that would be $(\epsilon,\delta)$-differentially private under bounded global sensitivity automatically achieves $(\epsilon,\delta,\gamma)$-random differential privacy (Hall et al., 2012), without any target-specific calculations required. We demonstrate on worked example learners how our usable approach adopts a naturally-relaxed privacy guarantee, while achieving more accurate releases even for non-private functions that are black-box computer programs.

This AI will tell people when theyre likely to die -- and thats a good thing. Thats because scientists from the University of Adelaide in Australia have used deep learning technology to analyze the computerized tomography (CT) scans of patient organs, in what could one day serve as an early warning system to catch heart disease, cancer, and other diseases early so that intervention can take place. Using a dataset of historical CT scans, and excluding other predictive factors like age, the system developed by the team was able to predict whether patients would die within five years around 70 percent of the time. The work was described in an article published in the journal Scientific Reports. The goal of the research isn't really to predict death, but to produce a more accurate measurement of health, Dr. Luke Oakden-Rayner, a researcher on the project, told Digital Trends.

Alcides Moreno and his brother Edgar were window washers in New York City. The two Ecuadorian immigrants worked for City Wide Window Cleaning, suspended high above the congested streets, dragging wet squeegees across the acres of glass that make up the skyline of Manhattan. On 7 December 2007, the brothers took an elevator to the roof of Solow Tower, a 47-storey apartment building on the Upper East Side. Over 420,000 people die worldwide each year after falling. Falls are the second leading cause of death by injury, after car accidents. They stepped onto the 16-foot-long, three-foot-wide aluminium scaffolding designed to slowly lower them down the black glass of the building. But the anchors holding the 1,250-pound platform instead gave way, plunging it and them 472 feet to the alley below. Scientists studying falling are developing'safe landing responses' to help limit the damage from falls. If you are falling, first protect your head – 37 per cent of falls by elderly people in a study by Professor Stephen Robinovitch of Simon Fraser University in British Columbia, involved hitting their heads, particularly during falls forward. Fight trainers and parachute jump coaches encourage people to try not to fall straight forward or backward. The key is to roll, and try to let the fleshy side parts of your body absorb the impact.

In late February, the UN-Secretary General held a press conference, highlighting the risk of starvation in East Africa and the necessity to raise funds to address the emergency situations in Somalia and South Sudan. Drought has been back in these countries and their neighbours since 2016, leading to a huge current food crisis. While governments are trying to handle the situation, how could technology innovations help prevent starvation and improve agriculture management in the future? We met with Laurent Rivière, a French 30 years-old entrepreneur, who shared with us his view on the subject with a combination of engineer pragmatism and changemaker idealism . Founder and CEO at Sunbirds for two years, he explained to us how his "bird of the sun," his solar drone, is addressing the agriculture challenges of the 21st century.

Queensland-based Icon Group has announced plans to adopt IBM Watson for Oncology, an artificial intelligence (AI) and cognitive computing platform that will provide the organisation's oncologists with access to massive amounts of global research to help them better inform patient treatment plans. "Being in the cancer care sector, we've been watching with great interest the way that Watson for Oncology has been utilised in other markets," Cathie Reid, co-founder of Icon Group, told ZDNet. "There's an incredible amount of medical literature already available and that's growing at an exponential rate and being able to have that information available in a curated form at the touch of your fingertips is something that our physicians are very excited about.

First up, I want to remind everyone – deep learning has really only been around as an applied method since 2012. So we haven't even had five years to use this stuff in medicine, and us medical folks typically lag behind a bit. With that perspective some of these results are even more incredible, but we should acknowledge that this is just the beginning. I'm going to review each paper I think is evidence of breakthrough medical automation, or that adds something useful to the conversation. I'll describe the research, but spend time discussing a few key elements: The task – is it a clinical task?

The first part of President Trump's $1 trillion infrastructure plan took flight Monday when the president called for the privatization of the country's air traffic control system. "We're still stuck with an ancient, broken, antiquated, horrible system that doesn't work," Trump said of the current air traffic control organization run by the FAA. Flights are delayed and flying sucks, the president argued, and the best way to change that is to take the responsibility for managing airspace away from the bureaucrats at the Federal Aviation Administration and hand it over to a private nonprofit organization. Such a plan won't cost taxpayers a dime, he said, and will improve efficiency. This idea dates to the 1970s, but has never gotten anywhere.