德國發表這份戰略報告的時間點，就在中國國務院總理李強預定訪問柏林的前幾天。被問及這份報告傳達給北京什麼樣的訊息時，朔爾茨表示重點是中國在經濟繼續成長，與世界經濟關係的整合不應受到影響同時，我們必須納入考量

德國發表國家安全戰略報告 批評北京試圖重塑全球秩序

今日（14日）的《東張西望》講到內地一個地鐵站，不知道什麼原因，一道捲閘突從天而降，在附近的人群立即四散，但還有一名女子走避不及，不幸被捲閘壓中。 相中可以見到捲閘塌下來後，大批市民都在旁圍觀，雖然該名女子的傷勢並不嚴重，圍觀的人都被指應該要先拉開女子或捲閘，以防萬一。雖然事件發生在內地，但主持容羨媛亦提醒說：「成日都提大家出街要『東張西望』，唔單止東同西，南同北都要望埋。」而陳庭欣就說：「所以話我出街鍾意著波鞋同運動褲，有咩事可以跑出啲同避開佢嘛！」

東張西望｜地鐵捲閘忽從天降壓中女子 容羨媛提醒出街要提高警覺

在本周一（12日）在開案當日瀏覽多個音樂平台，包括iTunes、Apple Music、KKBOX、Spotify、YouTube等均有顯示《願榮光》搜尋結果，亦提供多個音樂版本，包括創作人「湯瑪仕與眾香港人」版本。當中更以購買歌曲的iTunes，排行十大排名榜更曾清一色是《願榮光》系列或相關的歌曲。 不過，今日iTunes已經《願榮光》下架，已購買下載的音樂則可以繼續收聽，同樣是Apple音樂軟件的Apple Music，亦於下午起無法搜尋大部份《願榮光》版本，原本5個版本包括原創「湯瑪仕與眾香港人」版、純音樂版、英文版等已消失，現時只餘下一個台灣樂團的版本。Spotify亦於傍晚6時許時將音樂下架，顯示「無法播放歌曲」，用戶已無法收聽。

《願榮光》於各串流平台近乎清零 KKBOX晚上同下架只餘台語版

搭車讓座問題常掀爭議，內地河南鄭州一名老翁坐公車（巴士）時，要求一名大學生讓座遭到拒絕後，氣憤打大學生4巴掌，大學生未有還手；待大學生下車後，待在車上的老翁突然心臟病發猝死，事後老翁家人竟要求該大學生賠償50萬人民幣（約55萬港元），最終經法院審理，認為大學生毋須對老人的死負擔任何責任，判家屬敗訴。事件引來網民熱議，認為判決還大學生公道。

逼讓座被拒 老翁怒摑大學生後猝死 家屬反索賠$55萬 判刑爽了

長期夜間血壓高會導致心臟、腦部及腎臟等重要器官衰竭，根據研究報告指出，夜間高血壓會增加3倍患腦中風或心肌梗塞而引發死亡的風險，而患上失智症的風險亦是正常人的1.64倍。

夜間高血壓｜中風心肌梗塞風險增3倍！夜間血壓6症狀10食物助改善

警方周一（12日）接獲一名女子報案，指她在葵涌和宜合道一快餐店內懷疑被一名男子用手提電話拍攝其裙底。警方翻查閉路電視記錄，葵涌分區特遣隊人員鎖定一名涉案男子，今日（14日）進行埋伏行動，在葵涌和宜合道一快餐店內拘捕一名64歲姓呂男子涉嫌「非法拍攝或觀察私密部位」。案件交由葵青警區刑事調查隊第二隊接手調查，被捕男子正被扣留調查。

涉葵涌快餐店偷拍女子裙底 六旬漢遇警埋伏被捕

歐洲第二大晶片製造商意法半導體（ST）與中國半導體龍頭企業三安光電於日前宣布將聯合投資，斥資32億美元在重慶建造一家碳化矽晶圓廠。合資目的是為了滿足中國國內對碳化矽與日俱增的需求，以支援中國電動汽車以及其他工業電力和能源市場。有評論指，儘管美國政府一再向盟友施加巨大壓力，要求從旁遏制中國經濟，但商業「共生共存」為大，意法半導體合資三安落腳重慶設廠，極具象徵意義。

意法合資三安32億美元建碳化矽晶圓廠 落腳重慶具象徵意義

香港金牛籃球隊是日於北京參與簽約儀式，正式加盟全國男子籃球聯賽（NBL），因此球隊將會以香港為主場，參戰NBL。現時球隊除了梁兆華、楊睿騏、許烜行、范浩雲、張峰碩5位香港球員外，亦簽下具CBA經驗的內地球員張昊、孫偉博、孫巍、畢繼光、李佳益。

【本地籃球】香港金牛確定出戰NBL 7月主場迎戰冠軍隊

由Tiger執導的《是日精選》找來Lokman飾演男主角，二人由隊友變成導演及演員身份，Tiger大讚Lokman令人放心：「當初點解搵佢就係因為放心！佢不嬲都係演技擔當，咁啱風格同感覺上都好融洽。」而Lokman補充表示：「同埋係相對熟啲，落到場嘅時候唔使諗點樣講會舒服啲，好簡短咁講完就會慳咗啲時間。」

Tiger首執導爆Jeremy為角色爆粗極緊張 AK化身「顧問」傳授技巧

國泰航空航務董事甘沛思（Chris Kempis）表示，計劃為二副機長提供額外機會，推進事業發展；同時HK Express亦可在短期內得到機師，形容是雙贏。HK Express行政總裁毛潔瓊表示，HK Express的班次已超越疫前水平，會繼續從香港及海外招聘機師，並認為新舉措令兩家航空公司和機師都能受惠。

國泰邀二副機長到HK Express升任副機長 業界：快運人工低難外聘

Hk01

hk01

In this article, I want to introduce you to something truly amazing—Perplexity AI. This AI tool has incredible potential, and I believe it can be a game-changer for many of you. I'll walk you through how to use Perplexity AI and highlight some of its great features. Introduction to Perplexity AI Perplexity AI is an advanced search engine that works with real-time data. Unlike traditional search engines, it pulls information from various sources and provides a comprehensive summary. Perplexity AI utilizes advanced Natural Language Processing (NLP) algorithms and machine learning models to understand and process user queries. It employs a combination of neural networks and data parsing techniques to generate accurate and relevant responses. In NLP, 'perplexity' measures how well a probability model predicts a sample. Lower perplexity indicates better predictive performance, which is a key metric for evaluating the effectiveness of language models used by Perplexity AI. Getting Started With Perplexity AI To get started, simply open the Perplexity AI website (https://www.perplexity.ai). You can immediately begin by asking your first prompt. It's incredibly user-friendly and works just like a search engine but with more powerful capabilities. Perplexity AI can be used for answering questions ranging from basic facts to complex queries, creating code, and summarizing content, all sourced from up-to-date information. Additionally, it allows in-depth topic exploration with its Copilot feature, organizes your research with Collections, and facilitates data interaction and web searches within the platform. Focus Functionality in Perplexity AI Focus functionality is a feature that lets users specify the context or domain for their search queries. By selecting a particular focus, users can narrow down the search results to be more relevant to their specific needs, whether it's academic research, content creation, or any other specialized task. Academic Research: Users can focus their searches on academic papers, journals, and research articles. This is ideal for students, researchers, and academics who need in-depth information and credible sources for their studies or projects.


Content Creation: Writers, bloggers, and content creators can use the focus functionality to find relevant information, statistics, and references for their articles, blogs, or social media posts. This helps in creating well-informed and accurate content.


Video Descriptions: For YouTubers and video creators, this feature can help in generating detailed and accurate descriptions, scripts, or summaries for their videos. It ensures that the content is engaging and informative.


Business and Market Analysis: Professionals in business and finance can use the focus functionality to gather market data, financial reports, and industry trends. This aids in making informed business decisions and strategic planning.


Healthcare: Medical professionals can focus their searches on medical journals, clinical trials, and treatment guidelines. This is useful for staying updated with the latest medical research and improving patient care. How to Write a Good Prompts To write good prompts for Perplexity AI, be clear and specific, use complete sentences, include relevant details, ask open-ended questions, avoid ambiguity, use contextual keywords, break down complex queries, specify the format, use follow-up questions, and refine your prompts based on responses. The Difference Between Free and Pro Plans Free Plan - Free Basic Search Capabilities: Perform searches and receive comprehensive summaries from various sources.
No Registration Required: Start using the service immediately without creating an account.
Limited Data Handling: Attach files like PDFs or images with limited processing capabilities.
Standard AI Models: Access basic AI models for general tasks. Pro Plan - 20 USD/month Advanced Search Capabilities: Get more detailed and specific information, including access to specialized data like medical research.
Year-Specific Searches: Specify the exact year for more precise data retrieval.
YouTube Integration: Search for information on YouTube, and receive relevant video suggestions with summaries.
Personal Data Handling: Upload personal data, such as images or documents, for the AI to analyze and explain.
Creating Collections: Organize searches into collections for easier management and access.
Selecting AI Models: Choose from a variety of AI models, including GPT-4, Claude, DALL-E Playground, and Stable Diffusion.
Image Generation: Generate images based on user prompts, useful for visual projects or presentations.
Enhanced Sharing Options: Share search results easily for collaborative projects or sharing findings with others.
Pro Mode Toggle: Unlock additional functionalities and more powerful search capabilities. Video Tutorial About Perplexity AI https://youtu.be/ZvvgfCbNWTs?si=H018Q33lcSgyWvD5&embedable=true Visit my YouTube Channel Conclusion Perplexity AI is a versatile and powerful tool that has helped me with a wide range of tasks, from academic research to content creation. I highly recommend giving it a try. If you decide to sign up for a Pro account, you can unlock additional functionalities and enhance your research and productivity. P.S. You can get a $10 discount on the Pro account by using the link in the video description. :) In this article, I want to introduce you to something truly amazing—Perplexity AI. This AI tool has incredible potential, and I believe it can be a game-changer for many of you. I'll walk you through how to use Perplexity AI and highlight some of its great features. Introduction to Perplexity AI Perplexity AI is an advanced search engine that works with real-time data. Unlike traditional search engines, it pulls information from various sources and provides a comprehensive summary. Perplexity AI utilizes advanced Natural Language Processing (NLP) algorithms and machine learning models to understand and process user queries. It employs a combination of neural networks and data parsing techniques to generate accurate and relevant responses. In NLP, 'perplexity' measures how well a probability model predicts a sample. Lower perplexity indicates better predictive performance, which is a key metric for evaluating the effectiveness of language models used by Perplexity AI. Getting Started With Perplexity AI To get started, simply open the Perplexity AI website (https://www.perplexity.a i ). You can immediately begin by asking your first prompt. It's incredibly user-friendly and works just like a search engine but with more powerful capabilities. i Perplexity AI can be used for answering questions ranging from basic facts to complex queries, creating code, and summarizing content, all sourced from up-to-date information. Additionally, it allows in-depth topic exploration with its Copilot feature, organizes your research with Collections, and facilitates data interaction and web searches within the platform. Focus Functionality in Perplexity AI Focus functionality is a feature that lets users specify the context or domain for their search queries. By selecting a particular focus, users can narrow down the search results to be more relevant to their specific needs, whether it's academic research, content creation, or any other specialized task. Academic Research: Users can focus their searches on academic papers, journals, and research articles. This is ideal for students, researchers, and academics who need in-depth information and credible sources for their studies or projects. Content Creation: Writers, bloggers, and content creators can use the focus functionality to find relevant information, statistics, and references for their articles, blogs, or social media posts. This helps in creating well-informed and accurate content. Video Descriptions: For YouTubers and video creators, this feature can help in generating detailed and accurate descriptions, scripts, or summaries for their videos. It ensures that the content is engaging and informative. Business and Market Analysis: Professionals in business and finance can use the focus functionality to gather market data, financial reports, and industry trends. This aids in making informed business decisions and strategic planning. Healthcare: Medical professionals can focus their searches on medical journals, clinical trials, and treatment guidelines. This is useful for staying updated with the latest medical research and improving patient care. Academic Research: Users can focus their searches on academic papers, journals, and research articles. This is ideal for students, researchers, and academics who need in-depth information and credible sources for their studies or projects. Academic Research : Users can focus their searches on academic papers, journals, and research articles. This is ideal for students, researchers, and academics who need in-depth information and credible sources for their studies or projects. Academic Research Content Creation: Writers, bloggers, and content creators can use the focus functionality to find relevant information, statistics, and references for their articles, blogs, or social media posts. This helps in creating well-informed and accurate content. Content Creation : Writers, bloggers, and content creators can use the focus functionality to find relevant information, statistics, and references for their articles, blogs, or social media posts. This helps in creating well-informed and accurate content. Content Creation Video Descriptions: For YouTubers and video creators, this feature can help in generating detailed and accurate descriptions, scripts, or summaries for their videos. It ensures that the content is engaging and informative. Video Descriptions : For YouTubers and video creators, this feature can help in generating detailed and accurate descriptions, scripts, or summaries for their videos. It ensures that the content is engaging and informative. Video Descriptions Business and Market Analysis: Professionals in business and finance can use the focus functionality to gather market data, financial reports, and industry trends. This aids in making informed business decisions and strategic planning. Business and Market Analysis : Professionals in business and finance can use the focus functionality to gather market data, financial reports, and industry trends. This aids in making informed business decisions and strategic planning. Business and Market Analysis Healthcare: Medical professionals can focus their searches on medical journals, clinical trials, and treatment guidelines. This is useful for staying updated with the latest medical research and improving patient care. Healthcare : Medical professionals can focus their searches on medical journals, clinical trials, and treatment guidelines. This is useful for staying updated with the latest medical research and improving patient care. Healthcare How to Write a Good Prompts To write good prompts for Perplexity AI, be clear and specific, use complete sentences, include relevant details, ask open-ended questions, avoid ambiguity, use contextual keywords, break down complex queries, specify the format, use follow-up questions, and refine your prompts based on responses. The Difference Between Free and Pro Plans Free Plan - Free Basic Search Capabilities: Perform searches and receive comprehensive summaries from various sources. No Registration Required: Start using the service immediately without creating an account. Limited Data Handling: Attach files like PDFs or images with limited processing capabilities. Standard AI Models: Access basic AI models for general tasks. Basic Search Capabilities : Perform searches and receive comprehensive summaries from various sources. Basic Search Capabilities No Registration Required : Start using the service immediately without creating an account. No Registration Required Limited Data Handling : Attach files like PDFs or images with limited processing capabilities. Limited Data Handling Standard AI Models : Access basic AI models for general tasks. Standard AI Models Pro Plan - 20 USD/month Advanced Search Capabilities: Get more detailed and specific information, including access to specialized data like medical research. Year-Specific Searches: Specify the exact year for more precise data retrieval. YouTube Integration: Search for information on YouTube, and receive relevant video suggestions with summaries. Personal Data Handling: Upload personal data, such as images or documents, for the AI to analyze and explain. Creating Collections: Organize searches into collections for easier management and access. Selecting AI Models: Choose from a variety of AI models, including GPT-4, Claude, DALL-E Playground, and Stable Diffusion. Image Generation: Generate images based on user prompts, useful for visual projects or presentations. Enhanced Sharing Options: Share search results easily for collaborative projects or sharing findings with others. Pro Mode Toggle: Unlock additional functionalities and more powerful search capabilities. Advanced Search Capabilities : Get more detailed and specific information, including access to specialized data like medical research. Advanced Search Capabilities Year-Specific Searches : Specify the exact year for more precise data retrieval. Year-Specific Searches YouTube Integration : Search for information on YouTube, and receive relevant video suggestions with summaries. YouTube Integration Personal Data Handling : Upload personal data, such as images or documents, for the AI to analyze and explain. Personal Data Handling Creating Collections : Organize searches into collections for easier management and access. Creating Collections Selecting AI Models : Choose from a variety of AI models, including GPT-4, Claude, DALL-E Playground, and Stable Diffusion. Selecting AI Models Image Generation : Generate images based on user prompts, useful for visual projects or presentations. Image Generation Enhanced Sharing Options : Share search results easily for collaborative projects or sharing findings with others. Enhanced Sharing Options Pro Mode Toggle : Unlock additional functionalities and more powerful search capabilities. Pro Mode Toggle Video Tutorial About Perplexity AI https://youtu.be/ZvvgfCbNWTs?si=H018Q33lcSgyWvD5&embedable=true https://youtu.be/ZvvgfCbNWTs?si=H018Q33lcSgyWvD5&embedable=true Visit my YouTube Channel Visit my YouTube Channel Conclusion Perplexity AI is a versatile and powerful tool that has helped me with a wide range of tasks, from academic research to content creation. I highly recommend giving it a try. If you decide to sign up for a Pro account, you can unlock additional functionalities and enhance your research and productivity. P.S. You can get a $10 discount on the Pro account by using the link in the video description. :) P.S. You can get a $10 discount on the Pro account by using the link in the video description. :)

Mastering Perplexity AI: A Beginner's Guide to Getting Started

In Decentralization We Trust

Unprocessed image by on DocuSign Unsplash The “Maybe just a quick one” series title is inspired by my most common reply to “Fancy a drink?”, which may or may not end up in a long night. Likewise, these posts are intended to be short, but I get carried away sometimes, so apologies in advance. The PixelLib library is a library created by that provides easy out-of-the-box solutions to perform object segmentation with just a few lines of code. It supports both videos and images. PixelLib Ayoola Olafenwa There are two major types of segmentation that the library can help with: : Objects in an image with the same pixel values are segmented with the same colormaps. Semantic : Instances of the same object are segmented with different colormaps. Instance There is a variety of tasks that PixelLib can accomplish. In this article, we will go through some of them. Prerequisites Before starting, some necessary packages need to be installed: Tensorflow (version 2 and above), OpenCV, and  of course, PixelLib itself: pip install opencv-python
pip install tensorflow
pip install pixellib —upgrade Models For the   use cases  illustrated in this article, three models will be used: Mask R-CNN The h5 model file can be downloaded here Deeplabv3+ model trained on pascalvoc dataset PixelLib supports two deeplabv3+ models, Keras and TensorFlow. The Keras model is extracted from the TensorFlow model’s checkpoint. The TensorFlow model performs better than the Keras model extracted from its checkpoint. In this article, the Tensorflow model will be used. It can be downloaded . here Xception model trained on ade20k dataset The h5 model file can be downloaded . here Models and assets can be found in the repo page releases Everything is set up now. Let’s write some code, shall we? The folder structure I will be using is this: ├── app.py
├── input_images
│   ├── clem-onojeghuo-L_hK813fu9k-unsplash.jpg
│   ├── docusign-yiW2yzZNnFo-unsplash.jpg
│   └── jen-theodore-C6LzqZakyp4-unsplash.jpg
└── models
    ├── deeplabv3_xception65_ade20k.h5
    ├── mask_rcnn_coco.h5
    └── xception_pascalvoc.pb : The python script where all the coding will happen. app.py : Images that will be used for demonstration purposes. input_images : The saved models. models Import the necessary packages: app.py import pixellib
from pixellib.instance import instance_segmentation
from pixellib.semantic import semantic_segmentation
import cv2
from pixellib.tune_bg import alter_bg Image segmentation Let’s see how instance vs. semantic segmentation looks like applied on this image: Photo by on DocuSign Unsplash Instance segmentation using the Mask R-CNN model segment_image = instance_segmentation()
segment_image.load_model("models/mask_rcnn_coco.h5")
segment_image.segmentImage("input_images/docusign-yiW2yzZNnFo-unsplash.jpg", output_image_name="instance_seg.jpg",
                           text_size=8, box_thickness=5, text_thickness=5,
                           show_bboxes=True) : output_image Name under which the image will be saved. : show_bboxes Show bounding boxes. , , : text_size box_thickness text_thickness Size and thickness of the boxes and text. Semantic segmentation using the Xception model trained on ade20k dataset segment_image = semantic_segmentation()
segment_image.load_ade20k_model("models/deeplabv3_xception65_ade20k.h5")
segment_image.segmentAsAde20k("input_images/docusign-yiW2yzZNnFo-unsplash.jpg", output_image_name="semantic_seg.jpg") Notice how the semantically segmented image uses the same colormap for the same object types. Image tuning The underlying object segmentation capabilities of PixelLib can be applied to accomplish image tuning tasks as well. For example, we can change the background of an image and replace it with another one. I will use this image to serve as the foreground: Photo by on Jen Theodore Unsplash …and this as a background: Photo by on Clem Onojeghuo Unsplash The model used for this task is the Deeplabv3+ model trained on pascalvoc dataset. change_bg = alter_bg(model_type="pb")
change_bg.load_pascalvoc_model("models/xception_pascalvoc.pb")

#Change background
change_bg.change_bg_img(f_image_path="input_images/jen-theodore-C6LzqZakyp4-unsplash.jpg",
                        b_image_path="input_images/clem-onojeghuo-L_hK813fu9k-unsplash.jpg", output_image_name="new_img.jpg") and : f_image_path b_image_path the foreground and background image paths, respectively. : output_image_name Name under which the image will be saved. Let’s see the outcome: That is not too bad, is it? The little fella was teleported to a lovely sandy beach! Real-time image segmentation from camera capture I will apply the same semantic segmentation logic, using the Mask R-CNN model, but this time, the input source will be a camera capturing in real-time. There is a small amount of code that needs to be changed: capture = cv2.VideoCapture(0)
segment_video = instance_segmentation()
segment_video.load_model("models/mask_rcnn_coco.h5")
segment_video.process_camera(capture, frames_per_second= 15, output_video_name="output.mp4", show_frames=True,
                             show_bboxes=True,
                             frame_name="frame",
                             extract_segmented_objects=False,
                             save_extracted_objects=False) You might have noticed that I have included the and arguments here that are by default. If these are set to then the extracted objects will be saved as images. extract_segmented_objects save_extracted_objects False True And here are some screenshots taken from the processed video: PixelLib looks like it manages to confidently identify and segment some common objects (with a high probability). Final thoughts and next steps PixelLib is a library that promises to simplify object segmentation on images and videos with just a few lines of code, and it doesn’t fail on delivering. It is a good entry point to computer vision due to its simplicity and ease of use. If you are feeling adventurous, you can spend some time examining the source code. I would also recommend reading the paper, available on . It is effortless to read. Simplifying Object Segmentation with PixelLib Library, paperswithcode In this article, I covered some of the use cases and examples. There are, however, more examples in the documentation, such as using your own dataset to train the models. Feel free to have a look if you are up for it. Useful links PixelLib on Github Official documentation

PixelLib: Image and Video Segmentation [Maybe just a Quick One]

My better half teaches for the (IB) program at a nearby school. I had a previous encounter with their Math club, on the topic of “Math at work”. Back then, was focused on the roll-out of Scrum at scale, so I touched on Fibonacci numbers (used for ) and scratched the surface of queuing theory, , to model service time in a work queue. further mathematics International Baccalaureate work effort estimation M/M/1 queues in particular Fast forward a month ago, the further mathematics class completed a healthy introduction to Graph Theory, including Dijkstra’s algorithm and the traveling salesman problem. Students remembered the “Math at work” session and asked for a sequel on “Graphs at work”. Based on conversations at home, I fully expected wickedly smart kids to wander in that class, so this was bound to be fun — beyond the refreshing change of scenery. I published the code-like content to , including a where it made sense. Gitlab CD pipeline To warm up and gauge the audience, I started with an introduction to the usage of trees in general and file system trees in particular. From mounting IaaS data volumes to accessing secrets in Kubernetes, this is the kind of tree that IT and Software people keep climbing up and down all day long. You have to give it to the Math folks: they like clear concepts and unambiguous definitions — you can tell that they have been at it for a while. As says, Wikipedia A tree is an undirected graph that satisfies any of the following equivalent conditions: is and (contains no cycles). - G connected acyclic is acyclic, and a simple cycle is formed if any is added to . - G edge G is connected, but would become if any single edge is removed from . - G disconnected G is connected and the 3-vertex _K_3 is not a of . - G complete graph minor G - Any two vertices in can be connected by a unique . G simple path The Unix filesystem being a very academic example of a tree, a fuller system (such as an entire cloud) can be seen as a , where mounting operations allow users to stitch trees together. And bind-mounting operations let parts of the tree be transparently mirrored in convenient places. forest We imagined a data volume provided as a service on a cloud provider, containing the home, www and log directories. The OS of a virtual machine connected to the volume would identify it as a device under /dev, for example as /dev/xvdh on AWS EC2. Per , the data volume would be mounted under /media/vol. Getting the subtrees home, www, log back where we expect them in the filesystem tree is achieved by bind-mounting them, as shown in the figure below. FHS The resulting graph remains a tree, because home is accessible both under / and under /media/vol. However, interacting with the files and children directories under /home has exactly the same effects as interacting with them under /media/vol/home. The value of this model is that the piece of software running against this tree does not need to (nor wants to) know where all the bits and pieces of tree are coming from. Only that the files and directories it relies on will be there, always in the place. separately right The same argument holds in the world of containerization. A container sees its (typically ) filesystem as a tree with always the right structure, no matter where the container is actually running. Container orchestration services such as Kubernetes take care of mounting data in the tree. The current best practice is to use the same mechanism to inject in the containers, using the filesystem — knowing that Kubernetes handles these “secret” objects as such, and not as regular data volumes. namespaced volumes secrets I then to Google’s 20-year-old page ranking algorithm, and went on their now famous In order to measure the relative importance of web pages, we propose PageRank, a method for computing a ranking for every web page based on the graph of the web. There is undeniable elegance in the concept of ranking a web page based on the rank of pages linking to it. And quite some power, as this simple equation made sense of the web as we know it. The paper first defines as the number of links from a web page . The graph of the web is captured in a square matrix , where is not far from an adjacency matrix of sorts ( if there is an edge from to ). PageRank is kept in , a vector of ranks over the web pages. is a normalization factor, so that the total rank of all web pages is constant. The PageRank inventors showed that is an eigenvector of with eigenvalue c: . If we want to inject a bit of Physics in this: we can see PageRank as the vector that describes the modes of vibration of the web — and as its level. N_u u A A A_{u,v}=1/N_u u v R c R A R=cAR 1/c energy Even though there are probably a number of additional algorithms in place to tweak these equations in Google Search, we spent some time pondering the kind of pages that such an algorithm will favor. How to best describe these top pages? Are they the most… pages? Or the most… pages? Or is it just some kind of beauty contest, where the winners are the most pages, whatever resonates with people, no matter if the pages are popular because they are or, a lot more problematic, because they are the center of attention, for the best or worst reasons we can imagine? important truthful popular accurate Switching gear and getting lower in the IP stack, I graphs as used in IP networking in general, and in OSPF in particular. With the advent of in the Linux kernel, it is possible to operate a large number of independent routing tables in a single host, therefore making it easy to simulate relatively complex network behavior. I borrowed the from Edwin Cordeiro and fixed it for recent versions of Ubuntu (18.04 LTS) and Mininet. introduced namespacing very good work OSPF Network A single VM can simulate a decent network with several hosts and routers , with a very limited memory and CPU footprint. For a limited investment, we can enter the brave world of Link State Advertisement and Shortest Path First (as known as Dijkstra). (h010_*) (r010_*) We can connect to the Zebra daemon in random routers, and start by investigating how routing works for directly-connected hosts. $ sudo nsenter -n -t $(cat /tmp/zebra-r010_1.pid) telnet localhost zebra
Trying 127.0.0.1...
Connected to localhost.
Escape character is '^]'.

Hello, this is Quagga (version 1.2.4).
Copyright 1996-2005 Kunihiro Ishiguro, et al.

User Access Verification

Password:
r010_1> enable
Password:
r010_1# show ip route connected
Codes: K - kernel route, C - connected, S - static, R - RIP,
       O - OSPF, I - IS-IS, B - BGP, P - PIM, A - Babel, N - NHRP,
       > - selected route, * - FIB route

C>* 10.0.0.0/30 is directly connected, r010_1-eth4
C>* 10.0.0.4/30 is directly connected, r010_1-eth5
C>* 10.1.0.0/24 is directly connected, r010_1-eth3
C>* 10.10.0.1/32 is directly connected, lo
C>* 10.255.0.0/30 is directly connected, r010_1-eth1
C>* 10.255.0.16/30 is directly connected, r010_1-eth2
C>* 127.0.0.0/8 is directly connected, lo
r010_1# What OSPF allows is for each router to know how to reach each of the networks advertised, i.e. to compute the next router that should be used to reach a remote network. In the example below, r010_1 learned from OSPF that it should send packets destined to 10.2.0.0/24 (a network behind r010_2) via its neighbor 10.255.0.2 through its eth1 interface. r010_1# show ip route ospf
Codes: K - kernel route, C - connected, S - static, R - RIP,
       O - OSPF, I - IS-IS, B - BGP, P - PIM, A - Babel, N - NHRP,
       > - selected route, * - FIB route

O>* 10.2.0.0/24 [110/20] via 10.255.0.2, r010_1-eth1, 00:44:36
O>* 10.3.0.0/24 [110/20] via 10.255.0.2, r010_1-eth1, 00:44:36
O>* 10.4.0.0/24 [110/20] via 10.255.0.17, r010_1-eth2, 00:44:36
O>* 10.5.0.0/24 [110/20] via 10.255.0.17, r010_1-eth2, 00:44:41
O   10.10.0.1/32 [110/0] is directly connected, lo, 00:45:27
O>* 10.10.0.2/32 [110/10] via 10.255.0.2, r010_1-eth1, 00:44:37
O>* 10.10.0.3/32 [110/20] via 10.255.0.2, r010_1-eth1, 00:44:37
O>* 10.10.0.4/32 [110/20] via 10.255.0.17, r010_1-eth2, 00:44:37
O>* 10.10.0.5/32 [110/10] via 10.255.0.17, r010_1-eth2, 00:44:42
O   10.255.0.0/30 [110/10] is directly connected, r010_1-eth1, 00:44:47
O>* 10.255.0.4/30 [110/20] via 10.255.0.2, r010_1-eth1, 00:44:37
O>* 10.255.0.8/30 [110/30] via 10.255.0.2, r010_1-eth1, 00:44:37
  *                        via 10.255.0.17, r010_1-eth2, 00:44:37
O>* 10.255.0.12/30 [110/20] via 10.255.0.17, r010_1-eth2, 00:44:42
O   10.255.0.16/30 [110/10] is directly connected, r010_1-eth2, 00:44:42 We end up with end-to-end connectivity, a key property of IP networks that is never overvalued. Yes, IPv4 NAT, I am targeting you. We can then examine end-to-end paths across the network: mininet> h010_51 traceroute h010_31
traceroute to 10.3.0.2 (10.3.0.2), 30 hops max, 60 byte packets
 1  10.5.0.1 (10.5.0.1)  0.060 ms  0.005 ms  0.004 ms
 2  10.255.0.13 (10.255.0.13)  0.015 ms  0.005 ms  0.005 ms
 3  10.255.0.9 (10.255.0.9)  0.017 ms  0.007 ms  0.006 ms
 4  10.3.0.2 (10.3.0.2)  0.024 ms  0.008 ms  0.007 ms Welcome in the magic of distributed systems, where nodes eventually reach a consensus on how they should behave individually so that the overall system performs the desired end-to-end routing function. As much as Dijkstra’s algorithm allows to find the shortest path in a finite search space, I could not resist and introduced the problem of in AI planning. The (hey, another kind of extremely important tree) contains a collection of tools for working with planning domains. Its very good on-line editor lets us import the tower of Hanoi domain, which contains a single action: searching a plan .domains Internet TLD (define (domain hanoi)
 (:requirements :strips)
 (:predicates (clear ?x)
              (on ?x ?y)
              (smaller ?x ?y))

 (:action move
  :parameters (?disc ?from ?to)
  :precondition (and
        (smaller ?to ?disc) 
        (on ?disc ?from) 
        (clear ?disc) 
        (clear ?to))
  :effect (and
        (clear ?from) 
        (on ?disc ?to) 
        (not (on ?disc ?from)) 
        (not (clear ?to)))
 )
) The domain is expressed in a -based language called (Planning Domain Definition Language). LISP PDDL The domain tells us that the objects we manipulate can be clear (they have no object on top of them), they can be placed on of top one of the other and that they are ordered via a smaller relation. The domain allows our world to be changed by moving discs, from a position to another one. There are four conditions to be allowed to move a disc from being on top of an object to being on top of another object: the disc needs to be smaller than the object it is moved to the disc needs to be currently on the object it is moving from the disc needs to be currently clear of any object the object it is moving to needs to be currently clear as well Solution to the tower of Hanoi problem with 4 discs (animation from Wolfram MathWorld) There are four effects of moving that disc: the disc we are moving from is now clear the disc is now on top of the disc we are moving to the disc is no longer on top of the disc we are moving from the disc we are moving to is no longer clear The problem definition lists all the objects we have in our world, along with their respective size and position. For example, with four discs: (define (problem hanoi-4)
 (:domain hanoi)
 (:objects peg1 peg2 peg3 d1 d2 d3 d4 )
 (:init
  (smaller peg1 d1)
  (smaller peg1 d2)
  (smaller peg1 d3)
  (smaller peg1 d4)
  (smaller peg2 d1)
  (smaller peg2 d2)
  (smaller peg2 d3)
  (smaller peg2 d4)
  (smaller peg3 d1)
  (smaller peg3 d2)
  (smaller peg3 d3)
  (smaller peg3 d4)
  (smaller d1 d1)
  (smaller d2 d1)
  (smaller d3 d1)
  (smaller d4 d1)
  (smaller d2 d2)
  (smaller d3 d2)
  (smaller d4 d2)
  (smaller d3 d3)
  (smaller d4 d3)
  (clear peg2)
  (clear peg3)
  (clear d1)
  (on d4 peg1)
  (on d3 d4)
  (on d2 d3)
  (on d1 d2)
 )
 (:goal
  (and 
   (on d4 peg3)
   (on d3 d4)
   (on d2 d3)
   (on d1 d2)
  )
 )
) Our four-discs problem starts by defining the three pegs and four discs, all of them being objects. It then goes on to define the initial state of the world. The pegs enter in the same order relation as discs: every disc is smaller than all pegs. Then d1 is smaller than all the discs included itself; d2 is smaller than itself, d3 and d4, d3. peg2 and peg3 are clear, while d4 is on peg1, d3 on d4, d2 on d3 and d1 on d2. The goal state is top have d4 on peg3, d3 on top of d4, d2 on d3 and d1 on d2. We can have a planner calculate the plan, i.e. the sequence of actions that will get us from the initial state to the goal state. (move d1 d2 peg2)
(move d2 d3 peg3)
(move d1 peg2 d2)
(move d3 d4 peg2)
(move d1 d2 d4)
(move d2 peg3 d3)
(move d1 d4 d2)
(move d4 peg1 peg3)
(move d1 d2 d4)
(move d2 d3 peg1)
(move d1 d4 d2)
(move d3 peg2 d4)
(move d1 d2 peg2)
(move d2 peg1 d3)
(move d1 peg2 d2) The planner invites us to move the disc d1 from being on top of d2 to being on top of peg2. Then, we should move the disk d2 from d3 to peg3, etc. Much of the power of this technology comes from its ability to replan if something goes wrong at some point during plan execution and we are left in-between the initial state and the goal state. The planner can reason based on a new initial state of the world, keep the previous goal, and devise a new plan that will let us reach the goal. , we can get the planner to show us the graph of the search space that it considered when looking for a solution. The links between the actions considered are called causal links. The plan, the sequence of actions that will get us from the initial state of the world to the goal state, can be derived without ambiguity from the set of causal links. By using a custom planner Search space We notice in particular that this planner starts by the goal state, and backtracks to the initial state. Beyond this toy example, AI planning is used to solve harder, real-world planning problems whose domain representation can include functions (floating-point numbers), require the planner to understand time or even reason about continuous change in functions. When the search state becomes large, visualizing it as a graph helps finding potential plateaus where the planner struggles to progress in its search. These plateaus can be eliminated by tweaking either the heuristic that the planner uses, or the domain definition. To cool down a bit, we with Directed Acyclic Graphs (DAG) as used in analytics workflows. Apache Airflow, or its Google-managed version Cloud Composer, allows to . Google’s example shows a very linear graph (create a Dataproc cluster, run a Hadoop job, delete the Dataproc cluster), so we expended it a bit: finished programmatically author, schedule and monitor workflows quickstart create_dataproc_cluster >> run_dataproc_hadoop >> delete_dataproc_cluster >> victory_bash
create_dataproc_cluster >> keep_busy >> victory_bash To a fairly large extent, the graph that we are specifying here manually as an input to Airflow can be compared to the output of an AI planner that we saw earlier. In our Airflow use case, the developer spells out the plan and Airflow executes it. In robotic terms, Airflow acts as a , i.e. executes the plan that some intelligence created to solve a problem. In the AI planning use case we saw earlier, the (human) domain modeler defines the individual actions (through their prerequisites and effects), the initial state of the world and its goal state, and the AI searches for the plan. See slide 12 of by Gerhard Wickler, U. Ediburgh to see where the two pieces fit in a robotic architecture. controller planner Automated Planning Getting back to our (robotic) : Airflow will draw the graph and walk through execution, specifying the execution status of each node. controller DAG for our simple workflow In order to be able to show the execution status, run after run, Airflow decomposes the DAG into a tree, walking backwards from the final operator: Tree representation Airflow can also represent the execution of a single run as a Gantt chart. The Gantt chart representation offers the ability to identify the critical path, i.e. the tasks that, if they were shorter, would reduce the total duration of the run. Gantt Chart Representation Full-fledged Gantt charts also typically specify constraints between tasks. They are relatively simple in the case of Airflow, with constraints in term of execution time and success of the preceding operator. In a robot, they can include invariants, failure conditions, all the way to operational constraints such as timeouts, minimum or maximum time between tasks starting up or ending. We reached the end of our quick through graphs. The mathematics were extremely light, but hopefully the students left convinced that graphs are an indispensable tool, in computer science as much as in Software Engineering or in Information Technology. <a href="https://medium.com/media/3c851dac986ab6dbb2d1aaa91205a8eb/href">https://medium.com/media/3c851dac986ab6dbb2d1aaa91205a8eb/href</a>

Graphs at Work. At school. And in other places, too.

by BlockEx 1. Chinese Govt-backed Fund To Invest in Japanese Yen-pegged Stablecoin Back in April, we on the Xiong’An Global Blockchain Innovation Fund, a Chinese government-backed blockchain fund. At the time, $1.6 billion was allocated to investments in startups. 30% of the amount was provided by the government itself. It has now been that Grandshores Technology wants to raise HK$100 million ($12.74 million), through a digital token fund, to issue a stable coin pegged to the Japanese yen. Grandshores Technology is indirectly controlled by Yongjie Yao, who is also a founder partner of the Xiong’An Global Blockchain. The funds, denominated in tether (US dollar-pegged stablecoin), will be raised from investors living outside of China. Yao believes that in the next three to five years blockchain will become mainstream and this is why he is currently focusing on the funding of blockchain and crypto startups/projects. reported reported 2. US Regulator Calls For ‘Do No Harm’ Approach U.S. Commodity Futures Trading Commission (CFTC) Chairman J. Christopher Giancarlo believes that when dealing with crypto. He backed his statement by mentioning that when the internet was born decades ago, it was able to flourish due to the government not implementing aggressive regulations. He’s now calling for the same to be done with crypto. Giancarlo shared his thoughts while interviewed by CNBC during the Singapore Summit, which took place last week. However, it should be noted that he’s not ignoring the risks of fraud and manipulation, problems which often arise with foreign exchanges and precious metals. In fact, the regulator said that, in that case, regulators need to maintain a strong approach. Still, in regards to policy making, he feels it’s important to be moderate and well informed when making decisions. regulators should adopt a ‘do no harm’ approach 3. American National Standards Institute To Discuss Blockchain At Next Forum The American National Standards Institute (ANSI) has on its website that at the next Legal Issues and Joint Member Forum, the World Standards Week 2018, the topics of discussion will be blockchain and artificial intelligence (AI). ANSI is an American “private non-profit organization that oversees the development of voluntary consensus standards” on various matters. At the event, there will be the possibility to , along with how blockchain could actually be employed. The list of attendees will also comprise of experts from various industries, including healthcare, transportation and food safety. ANSI previously set international standards for the C language, ANSI C, and the first computer programming language standard, FORTRAN 66. announced discuss ethical and legal concerns 4. Blockchain Can Help Societies Deal With Environmental Issues The World Economic Forum , a exploring how blockchain can help with the environmental issues our society has to deal with. The document was released last week at the Global Climate Action Summit. Blockchain-backed platforms could, for example, assist with the decentralisation of the management of natural resources such as energy and water. It could also lead to greater sustainability by improving supply chains. It is possible to affirm that the technology is already helping to bring more transparency. For example, at the beginning of this year, we the WWF’s decision to adopt blockchain technology to fight illegal fishing and tuna fishing slavery. How blockchain could save the planet was also . Here, it was discussed how it could bring more transparency to various industries. The same belief was reached in this report. PwC UK Partner Celine Herweijer noted that even if so far the focus has been on the financial services industry, blockchain could also help the environment. (WEF) created, in collaboration with services network PwC report covered one of the hot topics of the Blockchain Summit In Morocco This news roundup was brought to you by . BlockEx To receive our daily news roundup in your mailbox, sign up here: http://bit.ly/BlockExNewsAndUpdates

18/09/2018: Biggest Stories in the Cryptosphere

Improving your web page loading is not a trivial task. It takes a lot of time and requires a lot of knowledge. To make it easier we have library which should help you with that. pwmetrics How do you always measure? Let me guess — you open chrome, timeline tab and so on and so far. What do we provide you with? Command in console — . pwmetrics uses API for getting metrics values like , , , , and . pwmetrics Lighthouse [First Contentful Paint](https://docs.google.com/document/d/1kKGZO3qlBBVOSZTf-T8BOMETzk3bY15SC-jsMJWv4IE/mobilebasic?pref=2&pli=1) [First Meaningful Paint](https://developers.google.com/web/tools/lighthouse/audits/first-meaningful-paint) [First Interactive](https://docs.google.com/document/d/1GGiI9-7KeY3TPqS3YT271upUVimo-XiL5mwWorDUD4c/edit#heading=h.k1h25blerz3i), Perceptual Speed Index ([more about it](https://www.instartlogic.com/blog/perceptual-speed-index-psi-measuring-above-fold-visual-performance-web-pages)) First Visual Changing Visually Complete 100% Visually Complete 85% Those values you can get in format using json . pwmetrics https://example.com --json {"runs": [{"timings": [{"title": "First Contentful Paint","id": "ttfcp","timestamp": 82750857804,"timing": 838.263,"color": "green"}, {"title": "First Meaningful Paint","id": "ttfmp","timestamp": 82750857814,"timing": 838.273,"color": "green"}, {"title": "Perceptual Speed Index","id": "psi","timestamp": 82750873600,"timing": 854.059,"color": "blue"}, {"title": "First Visual Change","id": "fv","timestamp": 82750873541,"timing": 854,"color": "blue"}, {"title": "Visually Complete 85%","id": "vc85","timestamp": 82750873600,"timing": 854.059,"color": "blue"}, {"title": "Visually Complete 100%","id": "vc100","timestamp": 82750873541,"timing": 854,"color": "blue"}, {"title": "First Interactive (vBeta)","id": "ttfi","timestamp": 82750859471,"timing": 839.93,"color": "yellow"}, {"title": "Time to Consistently Interactive (vBeta)","id": "ttci","timestamp": 82750859471,"timing": 839.93,"color": "yellow"}],"timestamps": [{"title": "Navigation Start", "id": "navstart", "timestamp": 82750019541}],"generatedTime": "2017-05-17T11:50:08.096Z","lighthouseVersion": "2.1.0","initialUrl": "http://example.com","url": "http://example.com/"}]} Or you can use pwmetrics [http://example.com](http://example.com) to see fancy chart with , , , values. First Contentful Paint First Meaningful Paint Perceptual Speed Index Time to Interactive Results of run “ pwmetrics [http://example.com](http://example.com)" command Nexus 5X Note: Results are measured with device emulation, _5x_ CPU throttling, _1.6 Mbps_ download speed and _750 Kbps_ upload speed. To get approximated results use median run option. pwmetrics http://example.com --runs=3 It will get 3 different metrics and approximate them. Note: As far as _pwmetrics_ depends on _Lighthouse_ you can pass all flags to it, using CLI. Available Lighthouse flags - are described here . Example_:_ _pwmetrics --disableDeviceEmulation "chrome://history/" --save-assets_ CI usage Lets go further and decide to use it with CI. You really don’t wanna accept new changes which downgrade your site/app . We took care of it too. performance Run pwmetrics --expectations What do you need before running that, is to set up some config either in your or in your own file? package.json Other then , configuration file can be used. See more at package.json docs pwmetrics --config=./path-to-file.js package.json "pwmetrics": {  "url": "http://example.com/",  "expectations": {    "ttfmp": {      "warn": ">=500",      "error": ">=1000"    },    "ttfcp": {      "warn": ">=500",      "error": ">=1000"    }  }} Result of its work You also can look a we’ve made for integrating it with build . recipe tools Show results in a DevTools Timeline Viewer it’s standalone project helping developers share their traces between each other. DevTools Timeline Viewer How it works together: pwmetrics --view --config=./pwmetrics-config.js Read how to configure. docs https://goo.gl/4opmye What a background of this? We just get traces, upload them to your Google Drive and open Timeline Viewer with these traces. Ok. That’s it you might think. But not. We prepared one more cool feature. Analytics purposes It’s really helpful when you can submit metrics results into service/docs/sheets, make some graphics and share these results with others. Here it is pwmetrics --submit Read to configure config. more This command will submit all received metrics into . Google Sheets Note: For now we support just Google Sheets, but are open to new contribution to make it works with other services :) : Example pwmetrics --submit --config=./pwmetrics-config.js : Results See at . Google Sheets Enjoy! :) Thanks to ‏ ! @paul_irish @samccone @pedro93_ UPD: Article was updated due to version 3 release.

Easy Progressive Web Metrics

On Monday, YouTube kicked off an end-of-summer PR blitz, reporting that its Partner Program has . topped 2 million creators When you read about YouTube paying out $30 billion to creators over the last three years, and more than $3 billion in the second quarter of 2021 alone, you might be tempted to think that YouTube is the clear future of video platforms. I’d argue that these statistics distract from why centralized video platforms like YouTube aren’t good for creators or viewers. With centralized platforms, a single organization controls every aspect of the creator and viewer experience. For creators, this means that growing a following on a centralized platform is like doing home improvement in an apartment. They’re building up the value of something they have no stake in and no say in the operation of. Their followers can be taken away, and they’re subject to arbitrary rules enforcement and changes. Being a creator on YouTube specifically means supporting a platform that has for the types of creators that make it big. outright contempt Even for the creators who are able to make money from their content, they’re getting a lousy deal when it comes to revenue sharing on centralized platforms. Today, YouTube pays 55% of ad revenue to creators. So, over the last three years, you have a for-profit company that’s reaped more than $24B in advertising profit on the backs of creators' videos. Centralized video platforms aren’t any better for viewers. They use opaque algorithms that mean you have no idea what you’re being shown or why. Websites like YouTube often have their own agenda and will use their control to show you content that . they want you to see, whether you like it or not Decentralized, blockchain-based platforms for video solve many of the problems that creators and viewers can experience with the YouTubes of the world. In a decentralized model, creators own their channel — it's not possible for a channel to be taken away from the person who created it, similar to the properties of Bitcoin. Earnings on decentralized platforms settle near instantly, and are owned by the creators. Open-source code means the algorithms that return results cannot be . secretly manipulated For viewers, they don’t have to worry about their favorite creator being deplatformed. They also don’t have to be concerned about algorithmic manipulation that prioritizes what a centralized video platform would like the viewer to see — they know that the content that they’re shown is determined via an open, public algorithm, and have a real say in what appears on their homepage. One of the most common critiques of decentralized video platforms is that there isn’t a good way to deal with fringe content, and content in violation of Federal or State law. As someone that’s running a decentralized video platform that is now receiving more than 30 million monthly users, I’d say that these concerns are misplaced. Decentralized video platforms can put in place governance structures to deal with illegal content swiftly. But decentralized platforms will always seek to prioritize user choice, and not restrict content based on social or political beliefs. For a functioning democracy and progressing society, people must be free to exchange and debate ideas, even unpopular ones. All kinds of mainstream ideas, like evolution, heliocentrism, the treatment of ulcers, even democracy itself, were once fringe. Where would we be today if all once-uncommon ideas were immediately silenced? The future of video is decentralized. This model is faster, cheaper, and better for creators and viewers alike. Jeremy Kauffman is the CEO and founder of LBRY , a secure, open, and community-run digital marketplace built on blockchain. In 2020, LBRY launched Odysee , which is now the largest blockchain-based video platform in the world with several hundred thousand creators and 30 million monthly viewers.

Why YouTube’s biggest competition will come from decentralized video platforms

The takeover of high profile Twitter profiles last month on July 15 made headlines when public figures like former President Barack Obama, Jeff Bezos, and Elon Musk began announcing that they wanted to “give back” to the community. But instead of making a donation to a COVID-related charity or something similar, they were promising that if people would send them Bitcoin, then they would return twice as much as they were given. This type of social engineering scam is a common one. Bad actors will pretend to be someone famous and use a fake account to trick people into sending them money. In most cases, a closer inspection of the account will show that the account is fraudulent, usually by seeing that the user name isn’t the account that belongs to the person being impersonated. Avoiding these confusions over authenticity of an account is the reason why Twitter introduced their verified account status with the checkmarks next to the name. But this attack was different because the accounts really did belong to these high profile characters. If it had just been Elon Musk tweeting out something wacky, then the case might have been believable just like any other Wednesday. But the idea of Jeff Bezos giving back to the community? That sent most people’s radars blaring that something big might be afoot. (Image Source: Twitter) As it soon emerged that these were the real accounts tweeting out the scammy message, confusion and a lockdown quickly followed as Twitter went into crisis mode. Within hours, Twitter responded, shutting down tweeting services for all verified accounts, thus ending the scammers’ operation. Coinbase, the cryptocurrency exchange that was being used for running the ill-gotten gains halted payments to the illicit wallet, blocking digital currency from reaching the thieves’ hands. For a full rundown of the story, please see coverage by that even managed to get comments for their story. The New York Times from the scammers However, once control of the impacted accounts was regained, the questions over how the hackers had taken over these important profiles began. In from the company on July 30, Twitter claimed that the breach had been the result of a “phone spear phishing attack” that led to the hackers accessing their internal tools: a statement The social engineering that occurred on July 15, 2020, targeted a small number of employees through a phone spear phishing attack. A successful attack required the attackers to obtain access to both our internal network as well as  specific employee credentials that granted them access to our internal support tools. Not all of the employees that were initially targeted had permissions to use account management tools, but the attackers used their credentials to access our internal systems and gain information about our processes. This knowledge then enabled them to target additional employees who did have access to our account support tools. Using the credentials of employees with access to these tools, the attackers targeted 130 Twitter accounts, ultimately Tweeting from 45, accessing the DM inbox of 36, and downloading the Twitter Data of 7. (Source: Twitter) How did this Happen? In his of the incident, cybersecurity expert Graham Cluley notes that Twitter has not been particularly forthcoming regarding the details of the attack. He believes that the “phone spear phishing” — which is sometimes referred to as vishing, short for voice phishing — was likely the attacker pretending to be from the company’s Help Desk. He surmises that the hacker probably spoofed Twitter’s phone number and might have directed the victim to a fake login page where the credentials were captured. recent analysis While this is admittedly conjecture, his theory sounds solid given that the victim gave over their credentials to someone who was not on their team, which is most likely to happen with a Help Desk interaction. It is probable that the victim was directed to a fake login page where the capture occurred. Cluley goes on to note that these types of spear phishing attempts can “be even more successful when your staff are forced to work remotely because there’s a global pandemic.” But why is this so? Well, for starters, being on our own at home can make it much more difficult to confirm that the person on the other end of the line is who they say that they are. At the office, verifying that someone reaching out to you can be as easy as walking down the hallway or a flight of stairs. We can also turn to our co-workers and get a sanity check about whether something is or is not a good idea to do. Working from home is a different story. Lacking our regular support network, we need to increase our vigilance significantly. Don’t trust, and certainly verify multiple times. Mitigate Your Risk If we want to think about how organizations can help to prevent these types of attacks from being successful in the future, Cluley offers a useful bit of advice when he says that employees should be provided with the real Twitter support team long before they receive a scam call like this. He touches on two important points here. When you are approached to provide sensitive information in a way that might set off alarm bells, whether it be your boss asking you to pick up gift cards or change the payment details for a vendor, always confirm on a separate channel from the one that you were contacted on. By this we mean that if you received a sketchy email, then do not respond there but pick up the phone and speak to them yourself. The second is not to use the number or link that is sent to you. For example, do not use the support number that someone sends you supposedly from your service provider. Use the one from their official site that you are already familiar with. Also, do not click on links sent to you by mail to access your accounts. If your bank supposedly sends you an email that you have to urgently address an issue, then ignore their link and just log in on your own through their site. Zooming out a tick from the individual level, as organizations, we should ask ourselves if we are doing enough to reduce our risk. Are we providing the technical oversight to enforce security policies that an employee might unwittingly violate that can put the organization at risk? If a security event does occur, do we have the capability to quickly understand what happened and move to block additional damage? The question is not how to stop all human errors (errare humanum est), but how to reduce them and mitigate the damage when they do occur. As organizations, what are we doing to train our people how to spot threats with a healthy dose of paranoia but without going overboard? Versus Being Quick to Judge, Use This as a Lesson to Prevent Something Similar from Happening to You In reading about the details of the story and writing this post, the thought occurred that we often read the same advice about how to work securely. Tips like always making sure that your software is up to date, not reusing passwords, and of course confirming the identities of people contacting you for sensitive information all seem to be basic best practices at this point. It is hard to imagine that an attack as “basic” and cliche as impersonating the Help Desk could work. But the truth is that these types of social engineering tactics continue to provide fantastic results for hackers, and those numbers are hard to argue against. That’s why they continue to use them. Their victims aren’t dumb, but they are busy and stressed. To address these problems at their core and empower a new era of working from home and/or outside the regular traditional office setting - as - be sure to have education and protocols in place that are shared on a regular basis; best practices enacted to ensure the and effectively transparent to help mitigate the risk of insider threats that could result in customer data breaches, intellectual property theft, misuse of company assets or information through negligence or malfeasance. Twitter employees may be able to do indefinitely safety and security of remote workers employee monitoring of remote workers This article was originally published on IT Security Central and reprinted with permission.

Work From Home Accountable For Massive Twitter Hack

Although the PlayStation 5 is an excellent console in itself, it's PlayStation 5 DualSense controller makes the console stand out, having a range of features that make you feel like you are experiencing next-generation gaming. The next generation of video game consoles is upon us. Xbox has released the Xbox Series X and S, and the PlayStation 5 is the next instalment in the PlayStation series. Until recently, I was sceptical as to how these consoles are truly next-gen. The consoles' updated graphics and power seemed just iterations on previous hardware, not worthy of next-gen. Even when I picked up an Xbox Series S, I failed to see what made these new consoles next-gen worthy. However, that all changed when I managed to beat the scalpers and land a PlayStation 5. PS5 Controller's Haptic Feedback Rumble technology in video game controllers has been around for decades. In the new PS5 controller, rumble has been enhanced to a whole new level with haptic feedback. Haptic feedback uses dual actuators to replace traditional rumble motors, giving you immersive feedback when performing game actions. Haptic feedback is a standout feature on the PlayStation 5 controller, as it does genuinely immerse you in your gameplay. One example is its use in Astro's Playroom. Astro's Playroom is a fun 3D platformer designed to show off the PlayStation 5's capabilities. From just holding the controller, the haptic feedback technology made me feel every game environment, from my characters' footsteps to defeating enemies in the game's overworld. The DualSense controller gave me feedback on my actions, fully immersing me in the Astro Playroom world. Likewise, the haptic feedback technology works just as well when playing games such as Call of Duty: Warzone, giving accurate feedback when firing and shooting weapons. The use of haptic feedback makes this controller and console feel next-gen. Adaptive Triggers The PlayStation 5 controller also has adaptive triggers. Adaptive triggers, unlike typical standard triggers, offer a whole new level of game immersion using adaptive trigger technology to provide resistance and tension as you engage with your in-game environments. The use of adaptive triggers blew me away when I first experienced them during my playthrough of Astro's Playroom. In one particular instance, I found myself rock climbing up a cliff face. The adaptive trigger technology made me feel the tension and force behind every climbing motion and hold grab. Astro's Playroom makes full use of adaptive trigger technology. Likewise, although relatively few at this stage in the PS5's lifecycle, other games such as Call of Duty: Cold War are also making use of the adaptive triggers. In Call of Duty: Cold War, the adaptive triggers make you feel the tension behind every gunshot when you press down the trigger, which largely depends on the gun's weight. All in all, the adaptive triggers present in the PlayStation 5 controller are mind-blowing and fully immersive. Hopefully, in the future more games adopt the technology. A Built-In Microphone Unlike its predecessors, the PS5 controller has a built-in microphone, meaning you do not need to own a headset to communicate with your friends. If you are like me and constantly breaking headsets, this is a great feature. I was initially sceptical about the microphone's quality, but it is as practical and responsive as a voice chat feature as it is a voice command feature when using apps such as Youtube. It is also used in games such as Astro's Playroom and makes an excellent addition to the controller, making jumping on and playing with your friends on your brand new console instantaneous. A New and Improved PS5 Controller Share Button The PlayStation 5 controller also built upon its predecessor's share button. The share button on the PlayStation 4 allowed you to share screenshots and short clips with your friends. The new Share button on the Dualsense dubbed the create button does all this and so much more. The create button speeds up capturing screenshots and video sharing, making the process seamless. It also gives you the option to start broadcasting and streaming on streaming services such as Twitch. Likewise, the new button also allows you to choose what format you want to upload an image or video in. Thus, the new create button iterates on the PlayStation 4's share button to create a fluid and seamless sharing feature worthy of next-gen. 3D Audio The PS5 controller also supports 3D audio technology. For those who are not lucky enough to own a wireless 3D headset, the PlayStation 5 controller, through its headphone jack, supports a 3D Audio output with any headset. 3D audio is fully immersive, making you hear accurately at a distance every sound and audio cue and works well on the PlayStation 5, giving you the option to tailor the 3D audio to your preferences. Unlike Xbox, PlayStation supports 3D audio with all of your headsets, meaning anyone can pick up a controller and experience this immersion given they have a headset. It is an outstanding next-gen feature that anyone can experience thanks to the Dualsense. Final Thoughts on the PS5 Controller Overall, the PS5 DualSense controller is an outstanding piece of next-gen hardware. The PS5 controller sets the bar to what a next-generation controller can be and far outmatches its competitor's controllers. The PlayStation 5 controller provides a truly next-generation experience and is an incredible piece of technology. For more features on the PS5, including upcoming PS5 games, keep it locked on HackerNoon. Also, check out, New Warzone Bunker Locations: Find All the Bunkers in Season 2 5 Best Pokémon Fan Games You Can Play for Free in 2021 How to Get 12 Million Animal Crossing Bells via Shady Online Merchants

PS5 Controller Review: 5 Best Features That'll Blow Your Mind

Exploration versus exploitation is a topic that merits its own study in the field of AI and reinforcement learning. But it is far more than that. We consider this tradeoff every day in our lives, influencing every decision we make. A bit of background on what exploration versus exploitation is. Consider the graphic below. You are at the "X" in the lower left corner. The squares with borders are those that you can see. Your goal is to capture as many green rewards as possible. Red squares with a negative integer indicate a loss in reward. You have a choice. You can advance to the "5" square and end the game with a reward of 5. Or you can explore the environment and possibly finish with a reward of 49 (50 - 1 is the maximum). However, you risk losing reward, returning to the "5" square and ending the game with a reward of less than 5. The exploitation of knowledge we currently have is a task we constantly balance with the exploration of new knowledge. Do you start working after high school or go to college? Do you apply to a specialized program or explore various majors? Do you marry the first person you meet or date several? Do you hire the fourth person you interview or only after you've interviewed at least twenty others? Do you order the same steak and fries or try something new? These might seem either trivial or questions with existing, well-defined approaches. However, the danger lies when we make the exploration/exploitation tradeoff without being fully aware of it. A familiar example is that of starting a tech company. The reason why so many fail is often cited as, "No one used the product." However, the underlying reason was that the founders exploited their first vision of their business model and executed on that before sufficiently exploring the search space. They could have explored the environment by talking to customers, running A/B tests, validating the market, testing pricing - other techniques. While this would have consumed valuable time at the onset, it would have produced a much better decision algorithm that could have been exploited later, saving time in sum. Lean Startup I recommend a three-step process for reducing exploration costs and improving exploitation costs. Define your goal. Define your assumptions. Determine the levels of threat the assumptions pose to your goal and order them from weakest to strongest. Weak, high-threat assumptions should be explored first and turned into strong assumptions. If there are none of these, it might be time to increase the ratio of exploitation to exploration you pursue. This systematically de-risks your course of action and makes you well aware of the exploration/exploitation scale you have been implicitly balancing.

Stories, News & Mentions about HK01

Hk01’s Stories on HackerNoon

Hk01's News & Mentions