If AI (Artificial Intelligence) is the future, then AI plus Space equals future times two. According to Kenneth's Research, AI in Space Exploration market was valued at approximately USD $2 billion in 2018 and is anticipated to grow at a rate of more than 7.25% by 2026. This includes machine-learning solutions for detecting new planets, space weather using magnetosphere and atmosphere measurement, integration of AI in space vehicles and satellites, as well as AI-based robots that can perform highly complex tasks. In addition to Space Exploration, there is a more prominent market for AI in the Earth Observation (EO) sector, where according to the 10th edition of Euroconsult's report, the EO data and services market should reach USD 8.5 billion by 2026 based on current growth trajectories. This same report predicts a market value of USD 6.5 billion by 2026, considering only new applications and solutions that will be developed to open new markets.

Japan shouldn't ignore the data security risks posed by the Chinese video app TikTok, a senior ruling party official said. "Not only President Trump but also other countries such as the U.K. and India, are gradually becoming aware of the risks," Akira Amari, the ruling Liberal Democratic Party's tax panel chief, said Sunday on Fuji Television Network. "Since there are so many countries pointing out the risks, Japan cannot just stand by and watch." U.S. President Donald Trump on Friday ordered ByteDance Ltd., TikTok's Chinese owner, to sell its U.S. assets. Trump cited national security grounds, delivering the latest salvo in his standoff with Beijing.

The Human Rights Commission has warned that the federal government's growing reliance on artificial intelligence and automated decisions is dangerous and will increasingly put vulnerable Australians at risk. Releasing new research on Australians' attitudes on the use of artificial intelligence, Human Rights Commissioner Edward Santow has urged the government to overhaul its approach to the emerging technology to avoid another scandal like the "robo-debt" scheme that unlawfully calculated and pursued debts from welfare recipients. Government agencies are increasingly relying on automated decisions rather than humans. Mr Santow said the people that tended to be least aware of the rise of automated decisions by government agencies like the Australian Taxation Office and Centrelink were the most disenfranchised in society. "That is really, really concerning to me. And that's because there's a real tendency, when new technology is being trialled, essentially to beta test it on some of the most vulnerable people," he said.

I usually cover the developments in the field of Artificial Intelligence as part of my regular feature titled Tech Diaries, but this week was so jam-packed with AI news that I had to write a separate piece. One story that stood out from the rest was the AI-enabled photo-editing App called FaceApp. It is owned by Russia-based Wireless Lab and has been around since 2017, but with the recent addition of a feature that lets you see your future self made it go viral last week. My Facebook newsfeed was full of people showing off their'Now & Then' pictures -- Looked enticing but I held off the temptation to generate an older version of myself. For starters, some reports suggest that the App has collected more than 150 million photos of people's faces since its launch & according to its terms of service it can use the huge database in whatever way it wants.

Identifiable by blurry backgrounds and a never-changing eye-position, these profile pictures are not ones taken of real people, but instead created by a machine learning technology, thus allowing the group creating this AI-fueled campaign to pump out an army of convincingly-human social media accounts. General Adversarial Networks, the specific technology behind these manufactured profile pictures, create fake humans that can fool our own eyes by pitting two machine-learning algorithms against each other. One generates these faces while the other tries to spot if the face is genuine or AI-generated, till the face produced is almost impossible to tell apart from a real one, even by our own algorithms.

Facial recognition technology is now common in a growing number of places around the world from public CCTV cameras to biometric identification systems in airports already touching half of the global population on a regular basis. Visualizations from SurfShark classify 194 countries and regions based on the extent of surveillance. More recently, the Department of Homeland Security unveiled its "Biometric Exit" plan, which aims to use facial recognition technology on nearly all air travel passengers by 2023, to identify compliance with visa status. Perhaps surprisingly, 59% of Americans are actually in favour of implementing facial recognition technology, considering it acceptable for use in law enforcement according to a Pew Research survey. Yet, some cities such as San Francisco have pushed to ban surveillance, citing a stand against its potential abuse by the government. Facial recognition technology can potentially come in handy after a natural disaster.

The rate at which data is growing is astounding. Consider these statistics: the world's data will go from 33 zettabytes in 2019 to 175 zettabytes in 2025, according to IDC's "Data Age 2025" whitepaper. A study by Cisco estimates that IP traffic will reach 4.8 zettabytes by 2022, and Domo found that 2.5 quintillion bytes of data are created each day. Considering that the federal government is among the world's biggest generators of data, those numbers can be especially daunting. Government officials have worked to address those concerns.

While computer vision has become one of the most used technologies across the globe, computer vision models are not immune to threats. One of the reasons for this threat is the underlying lack of robustness of the models. Indrajit Kar, who is the Principal Solution Architect at Accenture, took through a talk at CVDC 2020 on how to make AI more resilient to attack. As Kar shared, AI has become the new target for attackers, and the instances of manipulation and adversaries have increased dramatically over the last few years. From companies such as Google and Tesla to startups are affected by adversarial attacks.

Recent advances in information extraction have motivated the automatic construction of huge Knowledge Graphs (KGs) by mining from large-scale text corpus. However, noisy facts are unavoidably introduced into KGs that could be caused by automatic extraction. To validate the correctness of facts (i.e., triplets) inside a KG, one possible approach is to map the triplets into vector representations by capturing the semantic meanings of facts. Although many representation learning approaches have been developed for knowledge graphs, these methods are not effective for validation. They usually assume that facts are correct, and thus may overfit noisy facts and fail to detect such facts. Towards effective KG validation, we propose to leverage an external human-curated KG as auxiliary information source to help detect the errors in a target KG. The external KG is built upon human-curated knowledge repositories and tends to have high precision. On the other hand, although the target KG built by information extraction from texts has low precision, it can cover new or domain-specific facts that are not in any human-curated repositories. To tackle this challenging task, we propose a cross-graph representation learning framework, i.e., CrossVal, which can leverage an external KG to validate the facts in the target KG efficiently. This is achieved by embedding triplets based on their semantic meanings, drawing cross-KG negative samples and estimating a confidence score for each triplet based on its degree of correctness. We evaluate the proposed framework on datasets across different domains. Experimental results show that the proposed framework achieves the best performance compared with the state-of-the-art methods on large-scale KGs.

We introduce a novel online multitask setting. In this setting each task is partitioned into a sequence of segments that is unknown to the learner. Associated with each segment is a hypothesis from some hypothesis class. We give algorithms that are designed to exploit the scenario where there are many such segments but significantly fewer associated hypotheses. We prove regret bounds that hold for any segmentation of the tasks and any association of hypotheses to the segments. In the single-task setting this is equivalent to switching with long-term memory in the sense of [Bousquet and Warmuth; 2003]. We provide an algorithm that predicts on each trial in time linear in the number of hypotheses when the hypothesis class is finite. We also consider infinite hypothesis classes from reproducing kernel Hilbert spaces for which we give an algorithm whose per trial time complexity is cubic in the number of cumulative trials. In the single-task special case this is the first example of an efficient regret-bounded switching algorithm with long-term memory for a non-parametric hypothesis class.