With headlines everywhere focusing on disposable plastics and air travel emissions, it's clear that our individual, everyday purchasing choices--from what we eat to how we get around--impact the world around us. But how about what we wear? According to the UN Alliance for Sustainable Fashion, apparel manufacturing produces 20% of the world's water waste and up to 10% of its carbon output and sends more than 21 billion tons of textiles to landfills each year. But it's also a $2.4 trillion dollar industry that employs more than 60 million people worldwide. Considering this scale and impact, the industry is at a crossroads, devising disruptive technologies, rethinking business models, and searching for innovation at every step -- design, production, distribution, and reuse.

According to research by WIRED and Element AI, a mere 12% of leading machine learning researchers are female. This gap also exists in industry. According to recent research by the World Economic Forum and LinkedIn, only 22% of jobs in artificial intelligence are held by women, with even fewer holding senior roles. The gap appears even more stark at the "FANG" companies--according to the AI Now Institute just 15% of AI research staff at Facebook and 10% at Google are women. The first step to bridge the gender gap in AI is awareness.

As artificial intelligence continues to mature, we are seeing a corresponding growth in sophistication for humanoid robots and the applications for digital human beings in many aspects of modern-day life. To help you see the possibilities, we have pulled together some of the best examples of humanoid robots and where you might see digital humans in your everyday life today. Even though the earliest form of humanoid was created by Leonardo Da Vinci in 1495 (a mechanical armored suit that could sit, stand and walk), today's humanoid robots are powered by artificial intelligence and can listen, talk, move and respond. They use sensors and actuators (motors that control movement) and have features that are modeled after human parts. Whether they are structurally similar to a male (called an Android) or a female (Gynoid), it's a challenge to create realistic robots that replicate human capabilities.

MUMBAI: Centre of Excellence (CoE) the country's first incubation centre for companies focussed on IP in Gaming, VFX, Artificial Intelligence, Computer Vision set up by Software Technology Park of India and backed by MeitY and The Government of Telangana was inaugurated on Monday. The CoE also signed memorandums of understanding (MoUs) with 5 partners TVAGA (Telangana VFX, Animation and Gaming Association), Hyderabad Angles, HYSEA, and IIIT- Hyderabad to provide resources like mentoring, technology support, infrastructure and funding. The start-ups will be mentored by a group of industry professionals and academics, led by the Chief Mentor, a Governing Council (GC) and Project Management Group (PMG). A total investment of 19.68 Crore will be spent over a period of 5 years in addition to existing infrastructure in the STPI facility. Start-ups on selectio will be assisted with a seed fund of Rs. 5 lakhs after due diligence by the PMG.

This article is based on a presentation given by Emerj CEO Daniel Faggella in Geneva, at the 2019 New Shape Forum: Weapons Governance for the Geneva Disarmament Platform. To learn more about Emerj's AI presentations and speaking, visit our presentations page. There are many more AI applications than there are governments and militaries experimenting with AI. But the AI products that are becoming more commonplace in the business world do have their use in national security and defense efforts. It will take years to update physical weapons systems, such as tanks, planes, and missiles, with machine learning technology, but there are ways to weaponize artificial intelligence without mounting it on any weapons.

The Cabinet on Tuesday approved a bill to support companies to develop secure 5G mobile networks and drone technologies amid growing alarm among Tokyo policy-makers over the increasing influence of China's 5G technology. The bill will give companies which develop such technologies access to low-interest rate loans from government-affiliated financial institutions if their plans fulfill standards on cyber security. Companies that adopt 5G technologies can also get tax incentives if they meet standards set by the government, according to the bill. The government will submit the bill to the parliament and aims to bring it to effect around summer. The United States has been waging a campaign against Huawei Technologies Co, which Washington has warned could spy on customers for Beijing.

DENVER – In this post-impeachment era of divisiveness and deadlock in the nation's capital, Uncle Sam has a message for top U.S. technologists: A Washington-based nerd strike force called the U.S. Digital Service is seeking private-sector coders, programmers and software engineers to make government user-friendly for a tech-savvy U.S. public. Launched after the 2013 crash of the Obama administration's Healthcare.gov website, the USDS recruits the nation's top tech talent for Peace Corps-style tours of duty to tackle the government's most pressing information management and online security problems. It has an increasingly rare distinction as an initiative supported by both the Obama and Trump administrations, according to current and former USDS staff and White House officials. "We've been enthusiastic about USDS since Day One," said Mathew Lira, a special assistant to Trump in the White House Office of American Innovation. Early USDS projects -- fixing the public-facing website of Obama's Affordable Care Act, helping green card holders apply for renewals electronically -- might not be top Trump administration priorities today. But many projects continue: Enabling electronic access to health records for millions of Medicare patients and their doctors; building a robust and navigable Veterans Administration website; securing civilian agency and Defense Department websites.

Recent study of adversarial attacks has revealed the vulnerability of modern deep learning models. That is, subtly crafted perturbations of the input can make a trained network with high accuracy produce arbitrary incorrect predictions, while maintain imperceptible to human vision system. In this paper, we introduce Block Switching (BS), a defense strategy against adversarial attacks based on stochasticity. BS replaces a block of model layers with multiple parallel channels, and the active channel is randomly assigned in the run time hence unpredictable to the adversary. We show empirically that BS leads to a more dispersed input gradient distribution and superior defense effectiveness compared with other stochastic defenses such as stochastic activation pruning (SAP). Compared to other defenses, BS is also characterized by the following features: (i) BS causes less test accuracy drop; (ii) BS is attack-independent and (iii) BS is compatible with other defenses and can be used jointly with others.

Despite the great achievements of the modern deep neural networks (DNNs), the vulnerability/robustness of state-of-the-art DNNs raises security concerns in many application domains requiring high reliability. Various adversarial attacks are proposed to sabotage the learning performance of DNN models. Among those, the black-box adversarial attack methods have received special attentions owing to their practicality and simplicity. Black-box attacks usually prefer less queries in order to maintain stealthy and low costs. However, most of the current black-box attack methods adopt the first-order gradient descent method, which may come with certain deficiencies such as relatively slow convergence and high sensitivity to hyper-parameter settings. In this paper, we propose a zeroth-order natural gradient descent (ZO-NGD) method to design the adversarial attacks, which incorporates the zeroth-order gradient estimation technique catering to the black-box attack scenario and the second-order natural gradient descent to achieve higher query efficiency. The empirical evaluations on image classification datasets demonstrate that ZO-NGD can obtain significantly lower model query complexities compared with state-of-the-art attack methods.

Calibration and equal error rates are fundamental conditions for algorithmic fairness that have been shown to conflict with each other, suggesting that they cannot be satisfied simultaneously. This paper shows that the two are in fact compatible and presents a method for reconciling them. In particular, we derive necessary and sufficient conditions for the existence of calibrated scores that yield classifications achieving equal error rates. We then present an algorithm that searches for the most informative score subject to both calibration and minimal error rate disparity. Applied empirically to credit lending, our algorithm provides a solution that is more fair and profitable than a common alternative that omits sensitive features.