For its NDR capabilities, the company was touted for its dedication to platform and innovation. Specifically, NetWitness’s full packet capture and detection with comprehensive network visibility stood out to GigaOm. As for its SIEM features, NetWitness ...

NetWitness Featured as Industry Leader in NDR and SIEM in 2023 GigaOm Radar Reports

The Texas Rangers and the Arizona Diamondbacks have provided an intriguing World Series matchup on the field. But on national TV, each of their low-rated games has had a bigger audience than the next

Holy Cow, Those Are Bad Numbers! World Series Games 1-3 Post Worst-Ever TV Audience Ratings

Synergies between our broadband and video businesses are now less compelling,’ CEO Patrick Harshman told investors

Harmonic Mulls Offers for Its Video Business

Meta, which owns Facebook and Instagram, told Newsbeat it had "taken action" against more than 675 million fake accounts between April and June this year. It said 98% of those were spotted before users reported them and the company works with police to investigate fraud.

Gig tickets: Online scams 'made me feel stupid'

In an 1,800-word Penske showbiz trade guest essay, the former Hulu and Warner chief implores media conglomerates to band together to form a bundled, ‘everything’ platform and refocus on building Netfl

Jason Kilar: Streaming Isn’t Broken, But Hollywood Media Company Streaming Strategies Sure Are

That second-tier bowl game your Power Five alma mater with the disappointing 8-4 record just accepted a bid to? It's now got the satellite TV company’s name on it

DirecTV Signs On To Become Title Sponsor of the Holiday Bowl

ARPU for high-speed internet improved, but Comcast lost 18,000 customers as T-Mobile and Verizon combined to add another 941,000 fixed wireless subscribers

Comcast Stock Drops As Cable Company Loses More Ground on Home Broadband in Q3, Sees Slowing Mobile Growth

A "woke-free" beer company raked in an astounding $500,000 ... Ultra Right Beer initially started as an online-only brand in April, but soon moved to stores and restaurants in southern states ...

Brew dubbed 'Conservative Dad's Revenge' released by Ultra Right Beer Company

An X post from CBS News describing how an Israeli family was "massacred" outside the Gaza Strip was strongly criticized on social media Thursday, leading the outlet to play cleanup online.

CBS News' tweet on Hamas attack on Israelis placing 'scare quotes' around word 'massacred' slammed online

With Verizon Communications reporting 384,000 FWA adds in Q3 on Tuesday, T-Mobile and Verizon now have more than 6.9 million fixed wireless customers between them. Notably, AT&T has also entered the arena, adding 25,000 customers in the third quarter.

T-Mobile Holds Steady With Fixed Wireless, Adds 557K Customers in Q3

GigaOM

gigaom

Cybercrimes are increasing at an accelerated rate. To put things into perspective, by 2025, the cost of cybercrime is predicted to hit a whopping $10.5 trillion annually. Statistics like these demand a significant need to enhance the data security posture in organizations across the globe. However, several obstacles affect the adoption of a robust data security framework, such as multi-cloud complexities, data and AI laws, and the limited capabilities of existing data detection and response, data loss prevention, and similar tools. Data Security Posture Management (DSPM) is a critical discipline that turns an organization’s focus from an infrastructure-first to a data-first approach to cyber security. This blog will discuss DSPM and the key capabilities and factors to consider when buying a solution for your data security needs. What is Data Security Posture Management (DSPM)? DSPM stands for Data Security Posture Management. Gartner coined the term DSPM a few years back in its Hype Cycle report. Gartner highlighted the need for enhanced data security measures or approaches to help prevent the growing instances of data breaches and overcome the complexities of multi-cloud that lead to those breaches. As mentioned earlier, DSPM is a data-centric approach to data security, be it for on-premises data stores, SaaS services, or hybrid multi-cloud environments. The solutions provide detailed insights into the existence and location of sensitive data across the environment, who accesses it or how it is accessed, and the security posture of the application or data stores where it resides. Core Capabilities of DSPM Organizations seeking the right DSPM solution to reinforce their data security strategy must recognize that not all solutions are created equal. In 2024, GigaOm released its first-ever assessment of DSPM in its Radar report. The research firm highlighted some key capabilities that must be fundamentally available in a robust DSPM solution. Let’s take a look at those capabilities. Data Discovery Data security starts with gaining visibility of all data assets across an organization, including its on-premise data stores, SaaS services, and other cloud environments, leaving no stone unturned. The DSPM solution helps an organization do that by automatically discovering its structured and unstructured data assets. Data Classification DSPM solutions automatically scan environments to identify and classify sensitive data based on its business value, regulatory requirements, and other industry-specific policies upon discovering data assets. Data classification is a critical process that helps organizations set appropriate controls and prioritize their most sensitive data. Data Flow Data movement across an organization's various systems, data stores, and applications is dynamic. Teams must visualize and understand the data flow across systems to evaluate its privacy and security controls. DSPM helps visually illustrate the data movement, enabling teams to trace the transformation of data and the associated risks. Risk Assessment Another critical feature of a DSPM solution is that it gives teams insights into the risks through a thorough assessment. DSPM continuously monitors the environments for various vulnerabilities, assigns risk ratings, and helps teams prioritize remediation. Data Access Governance One critical threat to sensitive data is excessive privileges that lead to unauthorized access or inadvertent exposure. DSPM leverages access insights and IAM integrations to help organizations gain visibility of data access and risks and set up appropriate access policies and controls based on users, roles, and permissions. It further paves the way to the principle of least privilege (PoLP). Security Controls Apart from access governance, DSPM further enables organizations to place appropriate security controls based on the classification of data and its relevant risk scores. Such controls may include data encryption at rest or in motion, dynamic data masking for secure sharing practices, etc. DSPM also helps automate the security controls across the environments to minimize the risk of human error. Data Lineage From data ingestion to retention, data transformation occurs continuously and rapidly. Tracking such transformation across its lifecycle becomes difficult when handling data at the petabyte scale. DSPM allows organizations to overcome this challenge by tracking the changes over time. Breach Response DSPM solutions are well-equipped to respond to data breaches as soon as they are detected. The tool leverages breach impact insights, such as the volume of data impacted, the affected identities, and the relevant regulatory requirements. These insights help the tool to automatically isolate the impacted systems, revoke access permissions on compromised data, and initiate the notification process. Enterprise Stack Integration DSPM shouldn't be just another siloed integration among other tools. In fact, it should offer a unified approach to data security by seamlessly integrating with the existing security stack. This key operational factor ensures that workflows remain seamless and the organization can maximize the security stack's value. Secure AI Like many other technologies, DSPM tools had to respond to AI's growing adoption and its unprecedented risks. The OWASP Top 10 for LLM Applications feature the top-most critical risks that could compromise the AI systems development lifecycle, resulting in legal fines, compliance violations, and reputational loss. Hence, advanced DSPM solutions must offer much-needed capabilities like data sanitization and redaction or LLM firewalls to help organizations safely adopt AI. Important Factors to Consider While keeping in mind the aforementioned capabilities, organizations should also consider the following factors when looking for a robust DSPM solution. Agentless Discovery It is imperative to ensure that the solution offers agentless data discovery across the organization's environment. This capability is necessary to enable fast data discovery and mapping while minimizing deployment complexities. Centralized Dashboard The tool should provide a centralized dashboard that aggregates various metrics and reports, allowing teams to simplify monitoring and provide real-time insights to stakeholders. Continuous Detection and Prioritization of Critical Data Exposure The solution should identify the most critical threats to the organization's security posture and implement immediate remediation controls. Data Lineage Mapping The tool should further offer data lineage capabilities, providing organizations with visibility into the transformation of their sensitive data over time. By tracking the data's lineage across its lifecycle, security teams can better understand the security gaps and thus establish optimized controls and policies. Automated Remediation DSPM tools should allow security teams to respond automatically to security incidents when they occur. This capability is essential in ensuring the integrity and confidentiality of the data. Scalability and Performance Lastly, organizations with hyper-scale environments should look for a DSPM tool that scales to accommodate the needs of their rapidly growing data. Cybercrimes are increasing at an accelerated rate. To put things into perspective, by 2025, the cost of cybercrime is predicted to hit a whopping $10.5 trillion annually. Statistics like these demand a significant need to enhance the data security posture in organizations across the globe. Cybercrimes are increasing at an accelerated rate. To put things into perspective, by 2025 , the cost of cybercrime is predicted to hit a whopping $10.5 trillion annually. Statistics like these demand a significant need to enhance the data security posture in organizations across the globe. 2025 2025 However, several obstacles affect the adoption of a robust data security framework, such as multi-cloud complexities, data and AI laws, and the limited capabilities of existing data detection and response, data loss prevention, and similar tools. However, several obstacles affect the adoption of a robust data security framework, such as multi-cloud complexities, data and AI laws, and the limited capabilities of existing data detection and response, data loss prevention, and similar tools. Data Security Posture Management ( DSPM ) is a critical discipline that turns an organization’s focus from an infrastructure-first to a data-first approach to cyber security. DSPM DSPM This blog will discuss DSPM and the key capabilities and factors to consider when buying a solution for your data security needs. What is Data Security Posture Management (DSPM)? DSPM stands for Data Security Posture Management . Gartner coined the term DSPM a few years back in its Hype Cycle report. Gartner highlighted the need for enhanced data security measures or approaches to help prevent the growing instances of data breaches and overcome the complexities of multi-cloud that lead to those breaches. Data Security Posture Management As mentioned earlier, DSPM is a data-centric approach to data security, be it for on-premises data stores , SaaS services, or hybrid multi-cloud environments. The solutions provide detailed insights into the existence and location of sensitive data across the environment, who accesses it or how it is accessed, and the security posture of the application or data stores where it resides. data stores data stores Core Capabilities of DSPM Organizations seeking the right DSPM solution to reinforce their data security strategy must recognize that not all solutions are created equal. In 2024, GigaOm released its first-ever assessment of DSPM in its Radar report . The research firm highlighted some key capabilities that must be fundamentally available in a robust DSPM solution. Let’s take a look at those capabilities. GigaOm released its first-ever assessment of DSPM in its Radar report Data Discovery Data security starts with gaining visibility of all data assets across an organization, including its on-premise data stores, SaaS services, and other cloud environments, leaving no stone unturned. The DSPM solution helps an organization do that by automatically discovering its structured and unstructured data assets. Data Classification DSPM solutions automatically scan environments to identify and classify sensitive data based on its business value, regulatory requirements, and other industry-specific policies upon discovering data assets. Data classification is a critical process that helps organizations set appropriate controls and prioritize their most sensitive data. Data Flow Data movement across an organization's various systems, data stores, and applications is dynamic. Teams must visualize and understand the data flow across systems to evaluate its privacy and security controls. DSPM helps visually illustrate the data movement, enabling teams to trace the transformation of data and the associated risks. Risk Assessment Another critical feature of a DSPM solution is that it gives teams insights into the risks through a thorough assessment. DSPM continuously monitors the environments for various vulnerabilities, assigns risk ratings, and helps teams prioritize remediation. Data Access Governance One critical threat to sensitive data is excessive privileges that lead to unauthorized access or inadvertent exposure. DSPM leverages access insights and IAM integrations to help organizations gain visibility of data access and risks and set up appropriate access policies and controls based on users, roles, and permissions. It further paves the way to the principle of least privilege (PoLP). Security Controls Apart from access governance, DSPM further enables organizations to place appropriate security controls based on the classification of data and its relevant risk scores. Such controls may include data encryption at rest or in motion, dynamic data masking for secure sharing practices, etc. DSPM also helps automate the security controls across the environments to minimize the risk of human error. Data Lineage From data ingestion to retention, data transformation occurs continuously and rapidly. Tracking such transformation across its lifecycle becomes difficult when handling data at the petabyte scale. DSPM allows organizations to overcome this challenge by tracking the changes over time. Breach Response DSPM solutions are well-equipped to respond to data breaches as soon as they are detected. The tool leverages breach impact insights, such as the volume of data impacted, the affected identities, and the relevant regulatory requirements. These insights help the tool to automatically isolate the impacted systems, revoke access permissions on compromised data, and initiate the notification process. Enterprise Stack Integration DSPM shouldn't be just another siloed integration among other tools. In fact, it should offer a unified approach to data security by seamlessly integrating with the existing security stack. This key operational factor ensures that workflows remain seamless and the organization can maximize the security stack's value. Secure AI Like many other technologies, DSPM tools had to respond to AI's growing adoption and its unprecedented risks. The OWASP Top 10 for LLM Applications feature the top-most critical risks that could compromise the AI systems development lifecycle, resulting in legal fines, compliance violations, and reputational loss. Hence, advanced DSPM solutions must offer much-needed capabilities like data sanitization and redaction or LLM firewalls to help organizations safely adopt AI. Important Factors to Consider While keeping in mind the aforementioned capabilities, organizations should also consider the following factors when looking for a robust DSPM solution. Agentless Discovery It is imperative to ensure that the solution offers agentless data discovery across the organization's environment. This capability is necessary to enable fast data discovery and mapping while minimizing deployment complexities. Centralized Dashboard The tool should provide a centralized dashboard that aggregates various metrics and reports, allowing teams to simplify monitoring and provide real-time insights to stakeholders. Continuous Detection and Prioritization of Critical Data Exposure The solution should identify the most critical threats to the organization's security posture and implement immediate remediation controls. Data Lineage Mapping The tool should further offer data lineage capabilities, providing organizations with visibility into the transformation of their sensitive data over time. By tracking the data's lineage across its lifecycle, security teams can better understand the security gaps and thus establish optimized controls and policies. Automated Remediation DSPM tools should allow security teams to respond automatically to security incidents when they occur. This capability is essential in ensuring the integrity and confidentiality of the data. Scalability and Performance Lastly, organizations with hyper-scale environments should look for a DSPM tool that scales to accommodate the needs of their rapidly growing data.

DSPM Buyer’s Guide: Key Factors to Consider

If you’re interested in my origin story with the founders at Sectorflow Inc., a startup building an AI platform for business, you can read more on this . LinkedIn post Here’s the quick version: I got a call from my former engineering team that they were quitting their jobs to start their own company and they wanted me on board as their COO. It was a hard yes from me — then I got pregnant soon after. Should I back out? I asked myself. I might be doing a lot of sitting around the next few months… “No problem. You can take all the time you need when the baby comes. We don’t envision doing this without you.” Cool - locked in - . As a first-time COO, I thought I should probably figure out what this role actually entails. I had over a decade of experience in , technical project management, , and business development but I had never served in this type of leadership capacity before. A search yielded several “First 100 Days as COO” articles with some solid advice, usually around setting goals and fostering relationships with the C-suite. Great stuff but I found it was pretty conventional whereas startup life and our team history is anything but that. Thus those well-meaning and unhelpful articles are the inspiration for this blog. I’ll discuss some of what I actually did in my first 100 days as a COO at an AI startup and cover tool selection, initiatives, research, and some general insights. anxiety subsides agile methodology product management Tool Selection The very first thing I was asked to do was pick and issue-tracking tools. With everyone in their own lane, communication was only happening via or the daily standup. I estimated that searching for information, duplicating tasks, miscommunications (or overcommunication trying to avoid miscommunication), and progress updates were costing us around 10 hours per week across the team, maybe more. Simply, it was time to get organized. project management Slack Our application for Notion For Startups had just been approved (this gives you 6 free months of the Notion Plus Plan) but we weren’t sure exactly how we would use it and the biggest question was, would use it? A quick pros and cons analysis led me to recommend using Notion for company matters and JIRA and Confluence for engineering ( ). And that’s almost exactly what we did. engineering see analysis below I started building out our first Notion workspace with the template, a good starting point if you’re going in fresh. Not every section was relevant to us but I left them all in there just in case. For example, we didn’t use the Fundraising Tracker table right away but a couple of months in - it was a ready-made interim CRM solution for our CEO. Overall, Notion is excellent for keeping track of non-engineering related. Data Room everything As for engineering and development, the team has experience with Atlassian tools so it was an easy ramp-up to creating sprints and managing tickets. What I didn’t expect was that we would completely disregard Confluence as our documentation tool and instead went with . ReadMe ReadMe is marketed as a doc hub for developers and our CEO (with an engineering background, of course) fell in love with it after a test drive. I admit I was concerned I wouldn’t be able to self-serve my way through some release notes but that wasn’t the case at all. I’m not a developer but I love ReadMe because it brings us value in several ways: So easy even a non-technical COO can use it Branded experience which keeps our doc hub consistent with our public image A dedicated location for us to nurture this key piece of our customer experience Internal accountability for the team, promotes discussion and decisions Sectorflow Tool Selection Analysis: Notion v. Atlassian Option Pros Cons Cost All-in-One with Notion Unified platform, user-friendly, highly customizable. Limited engineering-specific features, may not scale well. About $40-50/month for 5 members. JIRA & Confluence for Engineering, Notion for Other Areas Specialized tools for engineering, best of both worlds. Managing multiple platforms, potentially higher cost. Around $65/month (JIRA + Confluence) + Notion cost for 5 members. 💡 Key Considerations: Need for specialized engineering tools. Preference for unified vs. specialized platforms. Long-term scalability and integration. 📝 I recommend going with Option 2 (JIRA & Confluence for Engineering, Notion for other areas). My Recommendation: Here's why: Flexibility & Scalability: This option seems better suited for our growth and scaling needs. I'm anticipating the need for DevOps and sustaining engineering tools here. Distinct Project Management Needs: Development project management differs significantly from other business areas. Specialized Atlassian tools will better serve these requirements. Future Planning: Anticipating a PMO to handle project management down the line, having distinct tools will facilitate easier transition and separation of duties. Some final thoughts on tools. I’d like to give two other shoutouts here to for marketing automation and TrustPilot for SOC2 compliance. Like Notion, these SaaS companies approved us for their startup program. If you’re a startup trying to gain efficiencies on a budget, check if the tool you’re vetting offers a freebie before signing on. customer.io always Gotta hold on to that precious runway. Initiatives The most important thing I did in my first 100 days was lead the market research, ICP (Ideal Customer Profile), and messaging efforts for the startup. The reasons for doing these activities are easy to grasp - we wanted to engineer a sustainable growth and communication strategy for the company. What’s not so easy is the you have to do to get somewhere with your market positioning (there are levels to this). homework With no clear first step for market research, I had to work with what we have. Thankfully, my CEO was having tons of conversations with everyone we knew (existing investors, previous clients, and connections that the team had made over the years) so I gathered a list from him of all the contacts within our network. We then compiled this information into a table, along with some data about each contact's company. This initial research helped us identify consistencies or patterns, to give us some sort of insight into problems, trends, and use cases. did Another key piece to this process was engaging in numerous deep discussions with our founding team. We shared our knowledge and experiences, while also continuing to gather valuable information from the network. It raised several questions and uncertainties about our market fit and positioning. Were we an AIOps platform? A BI tool? To answer these questions, I took to Gartner, GigaOm, and Forrester reports to understand these industries better. By analyzing the players, their solutions, and the problems they were solving, we were able to determine that - well, we did not fit into any existing categories. Discussions continued. Research continued. This research and analysis process took several weeks, but ultimately, we developed a v1 ICP based on the research I did, the team’s limitations, and the team’s industry experience. In grad school, I learned about strategic communication plans and how they’re a great tool before putting out any communication on behalf of a company. It’s something I had done for companies before and intended to do for Sectorflow to get alignment on our messaging. However, as we were already in the process of identifying our ICP and having conversations about what we stood for and who we are, I came across and realized a is essentially the SaaS startup version of a strategic communication plan. This seemed a better framework to follow given where we were at as a young company. This was an easier part of the process for me, since waxing poetic about our philosophy reminds you why you’re doing all this crazy stuff in the first place. More importantly, the manifesto will now inform our go-to-market strategy and marketing efforts moving forward, setting the tone for our external communications. this video manifesto What I’ve learned Finally, I’ll leave you with the biggest lessons I’ve learned in my first 100 days as a COO: While you must understand what is going on in the product and engineering world, your main job is to be a business leader. Come in every day with an idea of what you should be doing and the expectation that you will not be working on that. As COO, I realized the need for extreme adaptability and pivoting based on what’s needed. You are the CEO’s support system. This is easier if you have a good relationship and thankfully, that’s the case at Sectorflow! Read. Learn. Repeat. Become a business student. Become an (AI)* student. It’s easy to go heads down in work (especially at a startup when there’s so much to do) but you can bring value just by knowing what’s going on. I carve out 1-2 hours per weekday for reading reports, news, and social posts to level up on artificial intelligence concepts and startup business practices [ Insert your industry here ] If you thought this peek into my first 100 days as a new COO was helpful, please connect with me on or follow as I continue sharing our startup journey. I welcome feedback and post ideas around fundraising, validation, or anything else you'd like to hear about. LinkedIn Sectorflow Additionally, I'm currently researching . If anyone works in AIOps/MLOps and wants to exchange ideas on responsible AI principles, as we build out the Sectorflow platform, I'd love to connect. 💙 AI explainability and ethics — never finished, Deborah

My First 100 Days As An AI Startup COO

এআই স্টার্টআপ সিওও হিসেবে আমার প্রথম 100 দিন

Meine ersten 100 Tage als COO eines KI-Startups

Mis primeros 100 días como director de operaciones de una startup de IA

Mes 100 premiers jours en tant que COO d'une startup d'IA

एआई स्टार्टअप सीओओ के रूप में मेरे पहले 100 दिन

AI スタートアップ COO としての最初の 100 日間

AI 스타트업 COO로서의 나의 첫 100일

Meus primeiros 100 dias como COO de startup de IA

Мои первые 100 дней в качестве операционного директора AI-стартапа

Bir Yapay Zeka Startup COO'su Olarak İlk 100 Günüm

100 ngày đầu tiên của tôi với tư cách là COO khởi nghiệp AI

我担任人工智能初创公司首席运营官的前 100 天

The Most Detailed Guide On MLOps: Part 1

MLOps-এর সবচেয়ে বিস্তারিত নির্দেশিকা: পার্ট 1

Der ausführlichste Leitfaden zu MLOps: Teil 1

La guía más detallada sobre MLOps: Parte 1

Le guide le plus détaillé sur MLOps : partie 1

एमएलओपीएस पर सबसे विस्तृत मार्गदर्शिका: भाग 1

MLOps に関する最も詳細なガイド: パート 1

MLOps에 대한 가장 자세한 가이드: 1부

O guia mais detalhado sobre MLOps: Parte 1

Самое подробное руководство по MLOps: Часть 1

MLOps Hakkında En Ayrıntılı Kılavuz: Bölüm 1

Hướng dẫn chi tiết nhất về MLOps: Phần 1

最详细的 MLOps 指南：第 1 部分

Artificial intelligence is a technology that surprised the world by evolving from a fictional tale in movies to now having a ubiquitous influence in every sphere of human life. But was this an unpredictable change? Not really. The algorithms and experiments started back in the 1950s. It built the foundation of the AI we know today, like business automation software, the infamous Alexa and one of the biggest accomplishments of AI this year - . ChatGPT This level of evolution is possible only if the right minds put on their thinking caps and begin execution in the right direction. Lets See The List of 10 AI Development Companies in USA 1. Google Cloud Google Cloud offers an array of machine learning and artificial intelligence tools for developers and data scientists. These tools can be used at every stage of the development pipeline, i.e. preparation, building, validation and deployment of AI platforms. Google Cloud’s services primarily include virtual resources and physical assets. Some tools available on the platform include , Vertex AI Workbench, Cloud Natural Language, Video AI, Vision AI, and AutoMl, to name a few. Its services extend to infrastructure development tools for building prototypes, scaling, and deploying projects. Vertex AI For instance, its Deep Learning Containers have pre-configured deep learning environments which can assist in building prototypes. And the high-performing GPUs help with scientific computing and ease specific workloads. For its remarkable vision and resources, in 2022, in Magic Quadrant for Cloud AI Developer Services. Some noteworthy clients of Google Cloud include PayPal, Gojek, 20th Century Fox, P&G and Twitter. Gartner named Google Cloud the leader Launched: April 7, 2008 Employees: 32000+ Stock Price: USD 96.94 (as of 16th February 2023) 2.  Inc. RisingMax Inc. RisingMax is one of the leading AI company, acts as a digital transformation catalyst for startups and FORTUNE-500 companies. They are obliged to assist businesses worldwide in reducing their carbon footprint and creating a sustainable future. With the strong eco-community, they support green tech initiatives and reward businesses working to build a better tomorrow. They celebrate the project's success by planting 100+ trees each time. By rewarding or compensating the plating cost, they also encourage our clients to plant at least 50 trees. 3. OpenAI Open AI is a non-profit AI research and deployment company based in San Francisco, California. Open AI’s hall of fame includes Chat GPT, DALL·E 2 and Whisper. Chat GPT is a masterpiece of conversational AI with the ability to produce unique dialogues based on inputs. DALL·E 2 can create original and realistic image compositions based on descriptions. Whisper is an automatic speech recognition tool that can go through any quality of audio in any language and transcribe it accurately. Additionally, Open AI has other development tools to offer, like GPT3 and Codex. Users can also leverage Open AI’s free content filtering, end-user monitoring and scoping of API usage services. Pacific Northwest National Laboratory, Scale Media Inc., and Stellar Science Ltd. currently use OpenAI services. As per , OpenAI aims to achieve revenue of $1 billion by 2024. With investors like Microsoft, Khosla Ventures and Reid Hoffman’s charitable foundation, the future of  AI innovation in OpenAI looks bright. Reuters Launched: December 10, 2015 Employees: 375+ Stock Price: Open AI is a private company as of February 16, 2023 4. IBM Cloud IBM is leading the AI sector with its AI technologies. IBM cloud features a list of tools, including AI and Machine learning tools, specifically . It was released in April 2010 and is still actively used to build machine learning models. Watson AI IBM Watson Assistant, Discovery, Knowledge Catalog, Language Translator, Knowledge Studio and Natural Language Classifier are some useful AI tools available on IBM Cloud. In 2020 AI market share report with the largest market share of 13.7%. IBM received the #1 ranking in the IDC IBM Cloud’s Watson AI is used by 70% of banking institutions and by some of the top companies like Panasonic, Covéa, DB Dialog & DB Systel and Humana Inc. Its use cases in multiple industries spread across 21 verticals, including insurance, aerospace, defense, banking and transportation. Launched: April 2010 Employees: 300k+ Stock Price: USD 136.4 (as of 16th February 2023) 4. Amazon Web Services (AWS) Amazon Web Services has a comprehensive set of AI and machine learning tools targeting business analysts, data scientists, and engineers, allowing them to build, train, and deploy their ML models even without ML experience. Users can engage with ML experts and utilize AWS’s consulting services to complete their projects. The platform has a list of free tier tools like Amazon Polly, , Amazon Comprehend, Amazon Translate, Amazon Lex and Amazon Transcribe. AWS boasts a partnership with industry leaders like Intel, Accenture, Deloitte and C3 AI. Amazon SageMaker AWS AI has been employed by many startups and enterprises, such as Amazon, Discovery, Intuit, Formula 1 and Cambia Health Solutions. For its successful operation, Gartner recognized it as a leader in the Cloud AI Developer services' Magic Quadrant. Launched: November 2016 Employees: 40,000+ Stock Price: USD 101.16 (as of 16th February 2023) 6. Alibaba Cloud Alibaba Cloud is one of the leaders in the cloud computing and AI industry, with tools and services for developers and enterprises. With its end-to-end solutions, such as data processing, model training, prediction and evaluation, it aims to make AI more accessible to all. Alibaba Cloud’s AI tools include technologies like Vision AI, Intelligent Speech, NLP, Interactive AI and Intelligent Marketing. Its services are spread across many verticals like finance, education, transportation and new retail. The recognized Alibaba Cloud as one of the top 3 public cloud enterprise container platforms. Companies like Newegg, Blackboard, PicsArt and Align Technology leverage the AI services of Alibaba Cloud. Forrester report Launched: September 2009 Employees: 3000+ Stock Price: USD 103.08 (As of February 16, 2023) 7. Microsoft Azure Microsoft Azure AI services create a new way to access breakthrough research in the industry. The platform’s easy API calls can help data scientists and data engineers access language, vision, speech and decision-making AI. Broadly, the Azure services include Azure Applied AI, Azure Machine Learning, Azure Infrastructure and Azure Cognitive Services, each of which can enhance business processes, accelerate performance and add an extra layer of security. Specifically, Azure offers machine learning tools like Jupyter, Notebooks and Visual Studio Code and open-source frameworks like TensorFlow and PyTorch. Besides its low-code and no-code tools, services give users access to powerful AI tools like GitHub Copilot and GPT3. Azure’s Open AI Microsoft Azure was recognized as a leader in the ‘2021 IDC MarketScape for worldwide general-purpose conversational AI platforms.’ GigaOm also positioned . Azure’s active clients include industry leaders like BBC, NHS, Twitter and NBA. Azure among the leaders in the MLOps Radar Launched: February 2010 Employees: 200,000+ Stock Price: USD 269.32(As of February 16, 2023) 8. Salesforce Headquartered in San Francisco, California, Salesforce presents a large infrastructure of AI tools for admins and developers to build, train and deploy AI projects. Einstein is the most well-known platform in Salesforce, with no-code pre-built solutions and custom AI features. It’s well known for laying a rich set of AI tools for smart apps, such as Einstein Prediction Builder, Einstein Bots, Einstein Vision, Einstein Language and Einstein Language Discovery. Salesforce has been recognized as a leader, for the 16th consecutive year, in the 2022 Quadrant™ for Sales Force Automation Platforms. Some of the well-known clients of Salesforce include notable names like Ford, Loreal, Hologic and Engie. Gartner® Magic Launched: 2016 Employees: 73000+ Stock Price: USD 171.10 (As of February 16, 2023) 9. Deepmind Started in 2010, Deepmind is a research laboratory with a team of scientists, engineers and ethicists innovating AI. It’s headquartered in London with offices in other parts of the world like Canada, France and the United States. , The company achieved success early on when its AlphaGo program became the first to beat the professional Go Player. After partnering with Google in 2014, Deepmind was able to develop programs that detect eye diseases. According to BBC Its innovative AI is used to personalize recommendations in Google Play and predict patient deterioration in the U.S. Department of Veterans Affairs. Launched: 2010 Employees: 1000+ Stock Price: USD 96.94 (As of February 16, 2023) 10. Samsung Samsung is bringing forth its research in AI to build innovative solutions and smart devices. Samsung’s AI is currently focused on AI Fundamentals, Language and Voice AI and Vision AI. The company has expanded to five countries and established 7 AI centers in Korea, the UK, Canada, Russia and the USA. At , the company unveiled its vision of innovating technology with AI-based semiconductors. Samsung became one of the first to create AI chips using Synopsys AI software. These AI chips, called the , are used in most of the Samsung smartphones built today, like the Galaxy S20 and the Note 10 Lite. Samsung AI Forum 2022 Exynos Today, the company’s list of smart devices also includes the AI EcoBubble washer, which analyzes the patterns and determines the best washing cycle for its users. Launched: March 1938 Employees: 266,000+ Stock Price: 63,700 KRW (As of February 16, 2023) 11. Neuralink Neuralink is a neurotechnology company based in the US and owned by Elon Musk. The company's purpose is to find the brain-computer connection and build a chip that can be implanted surgically in the skull. The chip will be called the N1 chip with an 8mm diameter and electrodes and insulation for the wires. In 2021, Neuralink made a showing a monkey with a brain chip playing a video game. According to , the chip may help bring back someone’s eyesight and fix any damage to the brain. video Musk Founded: 2016 Employees: 300+ Stock Price: USD 0.000008668 (As of February 16, 2023) Final Words: What Is the Future of AI? From AI software writing an entire novel to a chip that can control humans using an app, AI innovations are not easy to comprehend. But these companies have the resources, the skills and the determination to build something big in this direction. Our list of the Top AI development companies includes the leaders that are commanding the AI industry today. Soon, AI will take over every sphere of our lives and become an indispensable aspect of every business.

Top 10 AI Development Companies in USA

10 empresas de desarrollo de IA que lideran la innovación en IA en 2023

10 sociétés de développement de l'IA à la pointe de l'innovation en IA en 2023

10 एआई विकास कंपनियां 2023 में एआई इनोवेशन का नेतृत्व कर रही हैं

2023 年の AI イノベーションをリードする AI 開発企業 10 社

10 empresas de desenvolvimento de IA líderes em inovação de IA em 2023

10 công ty phát triển AI dẫn đầu về đổi mới AI năm 2023

2023年引领人工智能创新的10家人工智能开发公司

This story is a part of Hacker Noon's series of interviews. The series is intended for tech professionals contributing the most insightful Hacker Noon stories to share more about their writing habits, ideas, and professional background (and maybe a hobby or two). If you too would like to start contributing to Hacker Noon, . Meet the Writer you can do so here So let’s start! Tell us a bit about yourself. For example, name, profession, and personal interests. Wow, I can’t believe my cave-man writing has turned into things that other people enjoy! Where do we start…Hmmm…If we go far enough back in my history we can talk about growing up in a very rural area where it was quicker to drive to a Walmart, buy PC Gamer magazine (to get the free demo disc!) then it was to wait for the 28.8 Kbps dialup to pull anything down from the ‘net. My dad is the person who piqued my curiosity in computing, peering over his shoulder as he would battle the PC to get the games running.  He would later meet some smart guys who founded a very early search engine aggregator (shout out to the team!), it was these young guys that introduced me to C++ and (more importantly) the ability to take tech, creativity, and turn it into a career. Web Ferret My first job in tech was one in which I worked for free to get experience on my resume (I really did, I explore why in this YouTube video here: https://www.youtube.com/watch?v=5rrjVbvg5JY I worked there for three months and took my first paying tech job as desktop support in a hospital.  Since then I have worked in various roles from sysadmin, network engineer to security architect, and now CISO.  Every role has influenced my outlook on tech and I am thankful for the experiences I have had. Interesting! What was your latest Hackernoon Top story about? My last article was about , which is ‘Cryptos very first industry standard for security. CryptoCurrency Security Standard Do you usually write on similar topics? If not, what do you usually write about? While Crypto- sounds like it would be in the wheelhouse of a security professional, to be honest, my friends nor my colleagues ever expected me to put out something like this.  My bread and butter topics are things that keep CISO’s up at night, security engineering-focused content, and any wild story I can find in the security domain. anything Great! What is your usual writing routine like (if you have one?) As a security researcher for GigaOm, I have spent a significant amount of my recent history writing, every day.  As simple as it sounds, I found this to be the best thing for my writing skill set.  Just sitting down and committing to 1000 words (or some other arbitrary number) has been really helpful. I have also found that I think in complete sentences, but when I write something gets lost in the mix…and that’s ok!  I can always come back and easily add the missing pieces to make it sound like it did in my head. Of course, you have to be humble, accept your mistakes as learning opportunities, and take all feedback with grace. The editorial staff here, at HackerNoon, has been great as well! Being a writer in tech can be a challenge. It’s not often our main role, but an addition to another one. What is the biggest challenge you have when it comes to writing? Fighting imposter syndrome and finding the right words to use lol.  The latter is easily fixed. On one hand, I feel it is very hard to be original in the production of your content in the tech space because things can be very well documented (either by vendors or open source communities).  On the other hand, sometimes it takes unique understanding of technology and your method of explanation to help others learn and understand it. your What is the next thing you hope to achieve in your career? I want to usher the cybersecurity industry into a new era.  I know that sounds grandiose, but what I mean is still a lot of low-hanging fruit in the industry.  We just need to stop getting distracted by the “shiny new things” and look at what works and what doesn’t.  I haven’t yet figured out what it will take to make this happen, but I am working on it ;) Wow, that’s admirable. Now, something more casual: What is your guilty pleasure of choice? Hmmm… My wife and I have gotten pretty familiar with the restaurants in, and out of, the area (road trip for food) when we don’t feel like cooking!  I have been binging the Coin Bureaus videos on lately, but I don’t feel guilty about that. YouTube Do you have a non-tech-related hobby? If yes, what is it? Yes, sure do!  Naval Ravikant has a famous , turned podcast, where he explains his theory of how to achieve wealth and happiness. tweet storm In order to achieve wealth, which is choosing what you want to do when you want to do it (not having to “go to work” instead), you need to be sharp mentally and you need to maintain (or grow!) your cognitive capabilities. This requires physical exercise, regularly.  My son and I both play hockey.  I coach his team of “squirts” with two other dads a few times a week and I play one (late) night a week. What can the Hacker Noon community expect to read from you next? More crypto and more security, whatever that may be! Thanks for taking time to join our “Meet the writer” series. It was a pleasure!

'I Worked for Free to Get the Experience': Chris Ray, CISO & Security Researcher

Whenever the implications of Artificial Intelligence are discussed, the conversation tends to focus on high-income countries. Most of the research on this subject covers the United States and Europe while overwhelmingly developing world regions such as South Asia, Africa and South America are often overlooked. However, any discussion about automation that doesn’t include its impact on the developing world is myopic, to say the least. For more than two centuries, industrialization has been the key to poverty alleviation. Countries have achieved economic growth by setting up factories that employ a large number of semi-skilled workers. China has used this formula to lift 850 million people out of extreme poverty. But advanced AI promises to eliminate the need for labor-intensive factories, creating a dilemma for countries striving to achieve economic betterment. Understanding AI’s Impact on the Developing World It’s estimated that in agriculture. Compared to this, farmers and ranchers are just 1.3% of the American workforce. It’s also important to remember that India is the second most populous country in the world with 1.3 billion citizens. 42.39% of Indians work Most of the work associated with these sectors is routine and repetitive. It’s work that’s bound to get automated in the coming years. Taking this into consideration, automation will hit poorer countries harder and could lead to mass unemployment. Harvard University economist Dani Rodrik calls it where countries who have just begun to grow through industrialization are at the risk of losing their manufacturing jobs. “premature deindustrialization” “The term deindustrialization is used today to refer to the experience mainly of these
 advanced economies.” Rodrik writes in a paper published in 2015. “With some exceptions, confined largely to Asia, developing countries have experienced falling manufacturing shares in both employment and real value added, especially since the 1980s”. But even some of Asia's emerging economies are at risk of losing whatever they’ve achieved in the last few years. These countries are not yet rich enough to soften the blow automation will deal to them. For instance, Bangladesh owes its rise largely to textile. It’s the second biggest exporter of ready-made clothing after China and depends heavily on garments for foreign reserve. According to one estimate, 80% of textile jobs in the country could be lost when AI takes over. Preparing for the Incoming Disruption One thing is for certain, automation is coming. Manufacturers wouldn’t think twice about integrating technologies that promise to optimize their procedures while reducing cost at the same time. Sooner than later, the and this will create several challenges for countries such as Bangladesh. Naturally, the developing nations will aim to up-skill their population but that will be another mountain to climb. fashion industry will embrace AI Such efforts will require policy initiatives on a broad scale that will include everything from infrastructure development to social, healthcare and education provision. Even then, these will be insufficient considering the scale on which AI will disrupt industries. Developing countries have the option of disincentivizing automation by taxing the use of robots or incentivizing the use of human labor through tax breaks or cutting minimum wages. However, this could push companies into countries that don’t discourage automation —which is why it’s unlikely that low-income countries will ever go this way. AI Will Create New Jobs But for Other Countries Futurists such as GigaOM’s Byron Reese believe AI will create more jobs than it will displace. According to Reese; “AI will be the greatest job engine the world has ever seen.” While there’s no doubt algorithms would need human help for training and maintenance, and this will create new job opportunities —some believe the developing world might miss out. Ian Goldin, a professor at the University of Oxford, who has extensively researched globalization, will be concentrated in the developed world. “(The) fact that poor countries also tend to suffer shortages of highly skilled labour could further undermine their competitiveness.” he explains. believes most AI jobs Goldin believes the widespread use of 3D printing might make things worse for emerging economies. “The development of supplementary technologies will add to the challenge. 3D printing, will combine with AI to allow consumers in rich countries to manufacture individually customised clothes, shoes, devices and other products, by themselves, much closer to home.” Production of this sort will eliminate any need for companies to outsource their production to lower-income countries. Goldin believes this will lead to increased inequality and more concentration of wealth in parts of the world that are already rich. Another Strategy for Policy Makers Center for Global Development’s 2018 study advises policymakers in the developing world to stay ahead of the machines. “A—risky but potentially inevitable—long-term coping strategy for developing countries would be to anticipate automation trends and to try to (further) develop a productive post-industrial sector.” The authors point toward some Automation Resistant Sectors (ARS) that governments can invest in to create employment. “Such an ARS could, for example, involve the social, education and health-care sectors, and some forms of tourism, and infrastructure construction which are generally considered resilient despite increasing service automation”. However, the study highlights that these sectors do not provide high value addition as of it and may not sufficiently scalable. But it is one of the many ways low-income countries can save themselves from the onslaught of Artificial Intelligence. But There’s a Silver Lining AI-induced unemployment is one of the major challenges for the developing world. But the same technology promises to help these countries counter some of their long-standing issues. Children in remote regions of Africa often lack access to good educational facilities. Dapito, an adaptive learning platform helps students study remotely. It uses AI to design courses whose structure, assessment and content are based on the weakness and strengths of students. Some emerging nations are also short on doctors and have an extremely high cost for healthcare. AI could assist medical personnel and optimize the diagnostic process. Nirmai in India is doing exactly that by providing a cost-efficient way to detect breast cancer in patients. It uses Thermalytix, which is a computer-aided diagnostic engine enabled by AI, to provide a low-cost alternative to mammography. Likewise, developing regions also have to deal with a bulk of refugees due to political instability. It uses image categorization for training algorithms. In Uganda, Refunite is a microtasking platform that’s giving refugees small income for performing verified tasks. The Clock is Ticking Experts such as Ian Goldin argue that AI could improve health and education in poor countries and create new jobs for the people. However, they don’t believe there’s enough evidence to prove these technologies will disrupt the region in a good way. Even though there are examples in countries like Uganda, Kenya and India of AI changing life for the better, it isn’t certain that this could be replicated all around the developing world. Job security and inequality are the two biggest issues policymakers in emerging economies will have to address as the world embraces Artificial Intelligence. It’s high time for legislators in these countries to institute policies that will help them weather the storm.

How Automation Impacts Job Security and Inequality in Emerging Economics

The only thing constant is change, and the roles and responsibilities of the , and the entire IT department for that matter, has been shifting a lot over the last couple of years. EMC’s Ken Oestreich intelligently predicted this shift about five years ago with , discussing why thinks that the CIO would essentially become the new IT supply chain manager. He writes that “IT and its organization will gradually shift in the following domains,” then citing “functional roles, skill sets, consumption models, and operational functions. CIO his post for GigaOM he Oestreich was right. CIOs the new caretakers of the supply chain, and you can see it in each of the four areas he was writing about. Of course, it doesn’t take Sherlock Holmes to understand why — the digitization of just about everything means that IT departments and CIOs have taken on myriad roles and functions in their organizations. Nevertheless, not everybody has fully embraced this melding of IT operations with supply chain management yet, and might need a little help doing so. are Here are a couple of tips for CIOs IT departments in the beginning stages of implementation who wish to transition smoothly and successfully toward supply chain management. These strategies aren’t all-encompassing but they’ll give you a good jumping off point, and from there you can cut your own path. Networking Transitioning into the demands that the CIO adopts a style of networking that is both farther reaching and more succinct than before, but internally and externally facing. This is because internal IT supply chain management is akin to . role of “supply chain manager” 3PL, 4PL, and even 5PL supply chain logistics strategy Tweaking your internal networking strategies will allow for more frequent and efficient cross-departmental communication and collaboration, allowing you to effectively monitor the tech-status of the entire organization. Building rapport and capital with other departments means that when they have problems, they’ll have no qualms coming to you for support. Conversely, you’ll have an easier time with new adoption and implementation. technology External networking exists to supplement and support your internal networking and operations. Through your contacts you’ll pick up product knowledge and the latest in industry trends, and will be able to select from a wide variety of vendors and products integral to your organization’s operations. Internal Networking If you haven’t already, set up a couple of meetings with the head of each department you’ll be directly interacting with. Give yourself a little bit extra time in these initial meetings to address all concerns that those department heads might have, and you’ll gain much more in capital than you’ll lose in production. This means you might end up having a month that is dominated almost solely by meetings — but it will be worth it. Find out exactly what technologies and processes each department deems “mission critical” and reassure them that they will have everything they need to do their jobs. After recording these needs, wants, and expectations, exercise proper business courtesy and individually email the department heads you’ve met with detailing their specific requests as you understood them. Use this as a means to open up direct and ongoing communication lines with individual department heads, and encourage them to come to you first if there’s ever an issue or request. Let them know that you’ll take care of them to the best of your ability, prove it through action, and in return you’ll be given more willing and understanding responses to any changes — future or present — that your team will have to implement. External Networking Based on what you learned from the department heads, either try to integrate existing technology into new services, or find solutions that are familiar to cut down on integration downtime. Whether you have a vendor representative that visits you on-site or a contact that you speak to remotely, try to become acquainted with a single point-of-contact for each vendor or solution that you utilize. If you build a relationship with external entities, you’ll end up finding the best price as well as unique, tailored solutions. Make sure to touch base with these contacts every couple of weeks to stay informed of updates and new releases as well as any other information that might be pertinent to your operations. Last, make sure you pass along customer feedback from department users, positive or negative. Honest customer feedback is a vendor’s lifeblood, and they probably appreciate it more than you know. Staying One Step Ahead Networking both internally and externally will also allow you to direct technology choices and prepare the system for security issues. Building a Catalogue Take what you know about your internal departments and their needs, combine it with what the market offers, and create an easily updatable digital catalogue. Let each of your internal contacts know that the solutions offered within the catalogue have been inspected, vetted, and approved for use by the IT team. This is not to say that solutions outside of this catalogue are impossible to obtain, just that the catalogued solutions are deemed safe and integratable by you. If employees in any department wish to use any solution outside of this catalogue, a request must be submitted via the chain of command to IT. This gives employees clear choices and decreases the chance that your company will have to . deal with Shadow IT Building for Security Run a risk analysis report on each department to determine where risk is and to prevent security breaches. Using this, shore up your risk pain points — this means that you may need to build some new internal infrastructure, but it will be well-worth it. Make sure the technology is extremely relevant to your company and that you’ve communicated why it’s important. For example, security measures from acquisitions should follow PCI DSS compliance measures if they’re using customer data. “While internal security processes can be easily implemented and managed, external threats are more difficult… boost network configurations and VPN protocols for an added line of technology defense,” . “The availability of consumer data online (social media) has made it easy for hackers to build a convincing customer profile and then use it to commit fraud.” write the experts at Kombea Also, track employee use patterns and build your infrastructure around that. For example, innovations in connectivity mean that people using personal devices and working remotely are at an all time high. This means that your infrastructure must be securely built to accommodate telework. Integration Disruption and constant change have become commonplace in the Age of Apps, so constant integration is to be expected. This is where external networking can come in extremely handy, because, if you can keep your finger on the pulse of technological trends, security updates, etc., you’ll have a better chance of anticipating industry changes and acting upon them it in a timely manner. Once you know that changes are necessary, immediately invest time in preparing the system for the upgrade. Your main enemy when it comes to integration is downtime, and the more prepared you are, the less downtime you’ll experience. Transitions are tough as well, because learning new software doesn’t always come easy. The more you prepare, however, the better you’ll anticipate and successfully navigate hang-ups in each department. Consistent Orientation Even though it’s the last point on this list, it is by no means the least important. Transitioning your team to this “supply chain” mindset is absolutely critical to the success of the transition itself. Level with your team, and talk about how the changes are affecting you as a CIO, and how your day-to-day is changing as the IT department takes on more supply chain-oriented operations. After that, open up the floor for other employees to detail their experiences and take note of any shared production snags or negative trends. Team orientation will be an ongoing operation, and will probably creep its way into every IT meeting that you have while you’re transitioning and even after. Take note of where you’re at on the project each time you meet, and make sure that you’re not only still on schedule, but are still understanding of the mission and not getting sidetracked by frivolous details. Final Word This is not the end-all be-all on the topic, and these are just a few strategies that CIOs could implement. They’ll allow you to build a solid foundation toward shifting your role and your IT department’s role to that of IT supplier. They’ll shift your organization’s idea of what IT does, and in time will help you to launch new technologies or upgrades. These changes in industrial and business paradigm aren’t coming — they’re already here. Still, this doesn’t mean that we shouldn’t stay looking toward the future, because we should. Drones, for example, are . It’s only a matter of time before industries the world over begin employing them, and those that follow the above tips will be well-prepared to implement those changes when necessary. quickly looking like the future of shipping logistics The only choice now is to embrace the change or go the way of the buffalo while your organization and its members embrace the change without you.

How CIOs and IT Departments can Transition Toward Supply Chain Management Roles

Graph databases are the hottest thing around right now. Whether you are just getting started, or you are in one of the 51% of organizations already using them, this is the place to get your news and analysis. The popularity of graph databases has gone through the roof almost overnight it seems. Everything points this way: the , the , the response to my . trend lines from database engines reports from the Forresters of the world graph-related ZDNet posts Why is that, and should you care? Do you really need a graph database, and if yes, how do you choose one? That’s the million dollar question. Well, that’s more than one questions actually, and that’s more than a million in value there too. Either way, i can help you answer, starting today. This is the first edition of the monthly Year of the Graph newsletter. Every month i will collect, republish, and comment on the 10 most important Graph database related news items. Why me? Ever since i implemented my first graph database prototype in 2005, i have worked on award-winning research, consulted the (then) leading vendor on distributed query implementation, and lead teams of all sizes and shapes working with graph databases. I have also published a number of reports and articles with analysis on big data, distributed systems and analytics with and . Perhaps more importantly though: Gigaom ZDNet I work for nobody else but me, and by extension, you. No fluff, just stuff. No automatically harvested or sponsored posts. No vendor affiliations or hidden agendas. Just hand picked, curated content, and objective, concise analysis. Why the year of the graph? Glad you asked. In a nutshell, because graph databases are coming of age and getting attention. Some of it has to do with the infrastructure and the technology enabling graph to get mainstream. The use cases have always been there, and heavyweights like AWS and Microsoft are moving in this space. _It's official: graph databases are a thing. That's the consensus here on Big on Data among fellow contributors Andrew…_www.zdnet.com The year of the graph: Getting graphic, going native, reshaping the landscape | ZDNet 2. How do you model you graph? This is the question Dilyan Damyanov from Snowplow explores. Using graphs is part of what Snowplow does, with an emphasis on managing events. The analysis on different answers on how to model event data as a graph sheds some light on the fine art of modelling graphs. _In recent months we've been busy expanding the variety of storage targets available for Snowplow users to load Snowplow…_snowplowanalytics.com Building a model for event data as a graph - Snowplow 3. So, you have a Hadoop data lake. There’s graph data in there, and “graphy” queries you can do on that data. ArangoDB’s Max Neunhöffer gives examples of graph use cases, and provides an introduction to some graph algorithms you can use, as well as a walkthrough of how to get started with ArangoDB. _Graphs in the sense of linked data with vertices and edges are everywhere these days: social networks, computer…_www.oreilly.com Fishing for graphs in a Hadoop data lake 4. Graph is a much more natural and efficient paradigm for doing multiple joins (hops) than relational data. Furthermore, graphs also work well together with another mega-trend, machine learning. The inherent structure in graphs can be leveraged in your machine learning algorithms, as Graphistry’s Leo Meyerovich discusses with O’Reilly’s Ben Lorica. _The O'Reilly Data Show Podcast: Leo Meyerovich on building large-scale, interactive applications that enable visual…_www.oreilly.com Graphs as the front end for machine learning 5. In their original inception data lakes grant universal access to data in their native formats, yet lack the necessary metadata and semantic consistency for long term sustainability. What can give you that? The combination of enterprise-wide ontologies, taxonomies, and terminology, says Franz Inc’s Jan Aasman. _One of the most significant results of the big data era is the broadening diversity of data types required to solidify…_www.infoworld.com Harmonizing big data with an enterprise knowledge graph 6. Graph databases come in 2 main flavors. RDF is one of them, and Linked Data is a collection of related standards, including taxonomies, that enable graph-based navigation and querying. If you’re interested in the foundamendals of this approach, this account of the Network for Information and Knowledge Exchange’s workshop may be for you. _Conrad Taylor writes: The first NetIKX meeting of 2018, on 25 January, looked at new technologies and approaches to…_netikx.wordpress.com Making true connections in a complex world - Graph database technology and Linked Open Data 7. If you think a taxonomy sounds fancy, how about an ontology? As Teodora Petkova from Ontotext argues, this is not about a philosophical debate on the essence of being. In order to define what something “is” (to a computer program), information technology resorts to the use of ontologies. Some graph databases use those — this is how you build a knowledge graph. _A program that wants to compare or combine information across the two databases has to know that these two terms are…_ontotext.com Ontology, Meaning and Knowledge Management 8. The other graph database flavor is LPG — Labelled Property Graphs. One of the differences between RDF and LPG is support for taxonomies and ontologies, and as a consequence of this, inference: RDF has it, LPG does not. Or at least, that was the case up to now, says Thorsten Liebig from derivo GmbH. Their solution wants to support RDF-like reasoning on LPG. _Discover how the Neo4j graph database and the Cypher graph query language perform as an OWL reasoning engine when…_neo4j.com Neo4j: A Reasonable RDF Graph Database & Reasoning Engine [Community Post] 9. Titan, an open source LPG graph database, was an important piece of the graph database world. As Titan is no longer maintained, IBM and others have been supporting a group of people who stepped up and adopted its codebase, forking it as JanusGraph. Ted Wilmes from Expero is one of the architects of JanusGraph, and he recently spoke about the state of JanusGraph in 2018. 10. What is the current status quo in the graph database world? All the latest developments from AWS, Cambridge Semantics, Neo4j and Tigergraph, in the “ best overview of the graph database world(s) so far “, plus analysis on the options for querying graphs. _Fifty years ago, relational databases were neither ubiquitous nor standardized. The story of how we went from the…_www.zdnet.com Back to the future: Does graph database success hang on query language? | ZDNet Would you like to receive the latest Year of the Graph Newsletter in your inbox? Easy — just signup below. Have some news you think should be featured in an upcoming newsletter? Easy too — drop me a line here . Originally published at http://linkeddataorchestration.com/2018/04/03/yearofthegraph/

The Year of the Graph Newsletter: April 2018

Some things I learned while building a GPU rig for deep learning During this month’s , Apple said, paraphrasing, that they tried to put a in the 2013 Mac Pro, but it was too hot and caused the CPU to throttle. That was my interpretation at least. briefing on the Mac Pro TITAN X Apple has a tendency to put underpowered GPUs in their desktops, so it’s great to hear they’re seeking to improve this for the next Mac Pro. They’ve gotten plenty of that “pro” users want more powerful GPUs, such as NVIDIA’s or . feedback GTX 1080 1080 Ti by Pascal Eggert — [ ] 2017 Mac Pro concept Source Decades ago, CPUs added units (e.g. Intel’s ) for number crunching. Despite to improve use of the CPU, there’s no denying a recent away from the CPU to the GPU for this kind of computation. SIMD MMX techniques shift GPUs have become so powerful that we’re using them more, leading to a rise in general-purpose on GPUs. It’s a prominent time for GPUs now, AMD will be competing at the high-end again with their series, and NVIDIA is coming soon after that. computing Vega Volta “Most of the software out there that’s been written to target [certain kinds of high-end cinema production tasks] doesn’t know how to balance itself well across multiple GPUs but can scale across a single large GPU.”~ Craig Federighi It’s true don’t take much advantage of multiple GPUs, but the real point is that old weak GPUs won’t compensate for a new powerful one, and we shouldn’t beyond that context. It’s fairly likely that Apple will sell a Mac Pro configuration in 2019 containing two GPUs. some apps generalize So you need a fast computer A few years ago, if you were to buy a new Mac Pro, you could spend as much to upgrade the CPU ($3500) as it cost for the whole machine by itself ($3000). [ ] Source Earlier this month, Apple updated their Mac Pro configurations and pricing to something much more reasonable by today’s standards. A CPU upgrade from 6-core to 12-core is now merely : $2000 [ ] Source That cost matches the difference in Intel’s list price between the 6-core ( ) and 12-core ( ) processors, so Apple is simply passing along Intel’s suggested pricing. $580 $2600 Note that the clock speed of the 6-core unit is 30% faster than the 12-core one, so you’d be upgrading to slower cores. Apps that rely upon single-core performance will actually run slower on the more expensive CPU. Now consider that will buy you an NVIDIA . Last month this was the GPU available to the public. $700 GTX 1080 Ti most powerful For about the same cost as adding 6 CPU cores, you can have 3 of the most powerful consumer GPUs in the world. It’s a shame you won’t be able to plug all three of those monsters into the next Mac Pro. Assuming, that is, Apple designs it with similar expansion constraints as the cheese grater tower, which was the most expandable Mac Pro they’ve ever made. Two mid-range GPUs. Or a single high-end one. Apple supported dual GPUs before the 2013 trash can Mac Pro. In 2012, the cheese grater Mac Pro had a stock configuration that came with two AMD Radeon HD 5770 GPUs. A dual-GPU, modular Mac Pro (mid-2012) — Photo: / Tom Park CC BY-SA As with the GPUs in the subsequent 2013 Mac Pro, these were considered to be mid-range, not high-end. They are rated under 110-watts , so they draw less wattage and generate less heat than high-end GPUs (250 watts TDP is typical nowadays) of the sort that you’d want in a workstation-class tower. TDP “But you could replace those cards with more powerful ones,” one might say. While that’s possible, the box wasn’t designed to allow it. It has a decent power supply, . However, it provides only 150 watts to each video card, or a total of 300 watts. To support a pair of 250-watt GPUs, like the , you’d have to supply at least 200 watts more power somehow, possibly by wiring additional power cables from the drive bay ports or from the remaining two expansion slots. 980 watts GTX 1080 Ti Back in 2009–2012 when this Mac Pro model was being built, high-end GPUs tended to be 180–210 watts each, but even then you’d still be short at least 60 watts. In fact, Apple offered an alternative Mac Pro configuration with a single high-end AMD Radeon HD 5870, rated at - TDP. You couldn’t insert another one without additional custom wiring. The cheese grater case was simply not designed for a pair of high-end GPUs. 228 watts On the other hand, this Mac Pro had a replaceable CPU module that allowed upgrading the system from single to dual CPUs. That’s pretty cool — in most other computers, you’d have to replace the motherboard to upgrade to dual CPUs. But nowadays you’d want this kind of upgradeability with GPUs. OK, so let’s suppose Apple’s next Mac Pro is modular and expandable, like the cheese grater tower, and fitted with the latest tech: PCIe 4.0, Oculink-2, Thunderbolt 3, USB-C 3.1, M.2 NVMe SSDs, and LGA-2066 or possibly LGA-3647 sockets for high-end CPUs with 44 lanes of PCIe connectivity. That sounds promising, but how many GPUs will it support? Maybe two, if you fiddle with it? high-end Pfft, how many GPUs do you really need? This question reminds me of those people who were saying no one legitimately needs more than 16 GB RAM in a Macbook Pro. last year “Own the world’s greatest gaming computer and convince everyone it’s for your research” — http://imgur.com/wiLsGqA Three GPUs in a computer might seem ridiculous. Amongst the of Mac customers who buy a Mac Pro, an even smaller fraction would need more than two GPUs. single digit percentage There’d never be a controversy about Apple supporting only two GPUs. But if one came up, a person might claim that no one legitimately needs more than two GPUs. Then a person would be wrong. Two weeks, two weeks, two weeks If you try to make an AI program to recognize 1000 types of objects in a photo, one of the things you’d find out is that even with multiple GPUs it can take weeks or months to train a neural net. We can train a model from scratch to its best performance on a desktop with 8 NVIDIA Tesla K40s in about 2 weeks.~ Jon Shlens, Google Research The common advice to avoid this delay is to adapt one that’s already been trained, using a technique called “ ”. But depending on what you want to do, you can’t always be using someone else’s pre-trained neural net — at some point you’re stuck in a process where each iteration could take weeks. If at that point you’re using only one GPU, adding a couple more would reduce the turnaround time significantly. transfer learning Suddenly synthesized A common problem with neural nets created by “supervised training” is they need a huge amount of training data or else slight variations in input will fool them. That is, suppose you train a neural net to recognize cats, but your training data consists of photos where all the cats are sitting upright. It might not recognize a cat that’s upside down. Your program will handle these kinds of differences better if you augment the training set by adding rotated and resized copies of the originals. In a notable example, Baidu Research made a breakthrough in speech recognition, aided by layering noise over their initial data set. Baidu gathered about 7,000 hours of data on people speaking conversationally, and then synthesized a total of roughly 100,000 hours by fusing those files with files containing background noise.~ Derrick Harris, GigaOm But then they had 14 times more data, which takes that much longer to process. A few years ago Baidu . More recently they reported working with as many as or GPUs. used 8 GPUs 40 128 Now you have two problems Recently some of the coolest results in AI are based on a technique where two neural nets compete against each other (“ ”) to improve their ability to create and recognize types of data. generative adversarial networks With a larger model and dataset, Ian [Goodfellow] needed to parallelize the model across multiple GPUs. Each job would push multiple machines to 90% CPU and GPU utilization, but even then the model took many days to train.~ OpenAI Blog That means you’re training two neural nets, not just one. Again, you’ll want as much computing power as you can get. Would you pay extra for expansion slots? When I started doing deep learning work, I used GPU spot instances on Amazon cloud. After running up a small bill, I began looking into external GPU boxes, but ended up assembling a PC. The GPUs on AWS are now rather slow (one GTX 1080 is four times faster than a AWS GPU) and prices have shot up dramatically in the last months. It now again seems much more sensible to buy your own GPU.~ Tim Dettmers I bought just one GPU. But knowing that I might need to add more, I made sure that the PC would be able to take up to four. It’d been a long time since I’d built a computer from components, and way back then I wasn’t paying attention to how many GPUs I could put in it. To my surprise, many computers cannot support two GPUs at full bandwidth, and most cannot support more than two. There are only a few expensive motherboards that can support 4 GPUs simultaneously with x16 lanes of PCIe 3.0 connectivity. The ability to add 3 more GPUs cost me about $600. That’s about one-third of what the base system (w/o GPUs) would cost if built with standard components. This amount represents the higher prices for a quad-GPU capable motherboard, a big 1500-watt power supply, and a CPU with 40 PCI lanes (as opposed to 28 lanes). I literally paid extra for expandability itself. Would Apple sell additional expansion as an option? The Mac Pro is positioned as Apple’s most powerful computer, but computing power is coming in various forms. High-speed expansion slots provide the ability to upgrade computing power in whatever form it takes in the future. We’re not just talking about GPUs here — it might be , or like Google’s . Or . FPGAs ASICs TPU fast storage Computer manufacturers generally do not sell a single line of desktop computers with different cases. For example, you don’t usually see something like this: by Scott Richardson —[ ] 2013 Mac Pro concept Source Instead, the models of a computer line are positioned according to CPU speed, storage, and/or screen size. So this is what’s typical: https://www.apple.com/mac-pro/specs/ But doesn’t it make sense to sell optional expandability? It has value. You can put a price on it. C’mon you have to admit it’d be pretty cool 😎 I know what you’re thinking: “There’s no way Apple would do this. It’s not worthwhile because there aren’t enough customers for the bigger Mac Pro.” You’re probably right. And besides, it’ll have Thunderbolt-3, and maybe Oculink-2 which is even faster, so people can connect GPUs if they need them… you know, like this: external Photo: Peter Wiggins — [ ] Source Anyway, I’m looking forward to seeing how Apple rethinks the Mac Pro. When we hit three GPUs, we are technically in a niche category… From talks with ASUS, despite the fact that a product may be geared towards a niche market, that product may sell well to the standard market if it is perceived to be good.~ (review of multi-GPU boards) Ian Cutress, AnandTech Q&A Asking the tough questions. What’s the big deal about multiple GPUs — weren’t bitcoin miners putting 6 or 8 GPUs on a PC in a ? milk crate Yes, but cryptocurrency algorithms can run on a GPU without a lot of data transfer, so miners could use PCI riser cables or a PCI splitter, and connect each GPU using just one PCI lane. That means they didn’t need a 40-lane CPU and a quad-x16 slot motherboard. In bitcoin mining, GPUs were overtaken by ASICs. Won’t that happen in AI too? Probably, at least and are working on that. In the meantime, GPUs have versatility that make them useful for the future. Nervana Groq foreseeable ASICs have lower power requirements than GPUs, so wouldn’t a big power supply be unnecessary for multiple ASICs? Maybe, but then you might just get more ASICs on each expansion card. You mentioned FPGAs and ASICs, but what about quantum processors? Oh yeah, those too I guess. Why do you need 16 lanes per GPU? Games don’t show any difference between GPUs running on x8 vs x16. It’s true that games show no significant difference, but that’s probably because they’re tuned to run well on GPUs at x8. Games don’t saturate a x16 PCIe 3.0 connection, but other apps like deep learning programs can. How does a CPU with only 40 PCI lanes connect to 4 GPUs that each use 16 lanes? Right, 4×16 = 64, which is more than 40. The motherboard has two PCIe switches which each can multiplex 16 lanes of signal from the CPU to 32 lanes for two slots. So the 4 slots use only 32 lanes from the CPU. a similar kind of PCIe switch for the Thunderbolt 2 controllers in the 2013 Mac Pro. Apple used Don’t those PCI switches introduce a lot of latency? Some people say the latency is so bad that it’s better to just use 8 lanes, but that’s a . Apparently this idea spread due to game testing that showed slightly higher frame rates on x8 GPUs than on switched x16 GPUs. But games are not good for this kind of benchmark because they don’t exceed x8 bandwidth — you can get similar results on a unswitched x16 slot. myth I haven’t seen any good tests, but an NVIDIA showed very little latency or loss in bandwidth across a PLX (Avago/Broadcom) switch. benchmark Show us your rig. Pics or it didn’t happen. Here it is next to the cheese grater: An open case rig next to a Mac Pro 5,1 — Photo: / Tom Park CC BY-SA What the… ugly. Why no case? OK, so I have to vent a little about motherboard and case design. Cool desktop computer designs are totally thwarted by the way uATX, mITX, and ATX boards are standardized. This sector has so much room for innovation with non-standard motherboards and backplanes. A lot of people are calling for Apple to just put the next Mac Pro in a big box case, but I think it’d be a shame if that’s exactly what Apple does. They went to extremes with the 2013 Mac Pro but reverting to a standard tower box isn’t so appealing either. I wanted a smallish form factor but the motherboard is so big (extended ATX) that it would have to go in a fairly large case, and I didn’t like anything I saw. So I experimented with MakerBeam rails instead of imitating with a Carbide Air 540 case. It ended up being taller than I intended, but slim in other dimensions. NVIDIA’s DevBox A single-fan liquid cooler has been completely adequate for the CPU. The GPU is connected with a x16 riser cable and can slide all the way over the CPU, since the CPU doesn’t have a huge heat sink on top of it. That means multiple GPUs can be spaced apart with wide air gaps. It’s surprisingly quiet with two 140mm fans, even under sustained heavy load. What did make a huge amount of noise was a 5TB hard drive that was hammering away while feeding images to a convnet. After a couple annoying days of that, I switched to a SSD. It has no drive bays — there’s a M.2 drive and additional SSDs can be velcroed on a side rail.

The Mac Pro: A Case for Expansion

Twitter’s beauty is a well-designed constraint — broadcast what you want to say in 140 characters or less. No middlemen. The celebrity world flocked to Twitter. For the rest of us, a tweet sounded like a fable’s moral or couplets of yore*- short, simple and [hopefully] memorable. A simple approach that stood the test of time helped propel Twitter. During the recent earnings call, Twitter’s numbers spoke- 310 million users, 595 MUSD in revenue last quarter. Staunch users continue to be loyal. And Jack Dorsey and his Twitter team have done a remarkable job of keeping them happy. Yet, Jack Dorsey has a challenge. Wall street is baying for Twitter’s growth. Why is Twitter growth apparently stalled? Jack’s predicament reminds me of a story I watched with fascination in the villages in India. During my years growing up, every summer, I visited my grandparents in the village. One summer, I noticed something unique in the local mom and pop store in the village- I found sachets hanging visibly. I asked the shopkeeper what they were– he mentioned a local company made shampoo and conditioner in sachets [they looked like Taco Bell sauce packets, a little bigger] and the folks in the village found it convenient. When the city bred honchos created shampoos- they worked on better and deeper products — upping the ante every year in classier looking bottles. This local company made it into simpler nuggets and reached a larger customer base! By next summer, all the big brands replicated the model. The villager who sometimes masquerades as a storekeeper proudly shared his bigger display during my next summer visit. Back to the shampoo companies, who doesn’t want to grow their customer base! Could this village story have some parallels for Twitter? I wonder. Let us look at the big changes Twitter did last year- added a new timeline by algorithm vs. chronology, integrated periscope to timeline, added polls, added moments news channel and many more. It becomes apparent all changes, many overdue, made the product better. The simpler sachets are missing — the closest was the rebranding of favorites to a more generic “like” with a heart sign. It happens to the best of us — we get drawn deeper into what we know. The bigger question is: do the missing sachets matter? It only matters if there is a market outside your traditional sphere of influence. Let us look at the age demographics of Twitter vs. other social sites [thank you Pew Institute for your research of internet users in the US] Facebook is a behemoth — that is the obvious. Outside that, LinkedIn is relatively awesome with GenX and baby boomers and Twitter has ways to go in this demographic. In conversations with few of them [LinkedIn users who are more muted on Twitter], what dawned on me was how they viewed Twitter — conversations were hardly discussed. — a broadcasting platform with an emphasis on followers And a common Joe’s odds of a following are as good as or slightly better than his odds of starring in a Hollywood movie as the central figure. When the common Joes act in each other’s movies, that is a different story. Many a times, I have a hard time figuring out if I am hearing a sour grapes story or a genuine bewilderedness to Twitter style vernacular — #FF, MT, B4 and . One person shared with a chuckle — “Math has symbols and I was taught those when I was young, Twitter shortcuts is another story. I need the blessings of a shortcut goddess.” more Either way, a different type of simplicity is missing for these crowds. Making Twitter work for you, me and Twitter’s Investors What happens if Twitter is cast as a simple extension of browser bookmarks? As the world is moving into the cloud, could this be a natural extension without being a stretch? Is it a bit more appealing as a sachet to this crowd? I have been experimenting with Twitter for a while as a public bookmark and I found comfort. I liked two features — there is no expiry date to tweets [unlike updates on LinkedIn] and any fellow Joe can drop by and get a feel for my tweet bookmarks at their leisure without conversations hampering the flow of the list. The upside in all this — any advertiser can look into my bookmark tweets [they are public] and hopefully target better customized ads — if I cannot avoid the ads [I do realize that they help keep content free], let these ads, at least, be more relevant. More than anything, I use bulk of the 140 characters to remind myself why I bookmarked the article as a tweet. Have I conquered Twitter? Heck, no. Gradually, I figured out the merits of hashtags and @mentions in my tweet bookmarks. After the fact, I wonder why I did not have the fortitude to make my twitter handle ultra small. I am really envious of Om Malik, the founder of GigaOM (awesome site that is unfortunately erstwhile). He has one of the shortest- @om. [Jack- maybe, you can excuse the long handles from the 140 counts to encourage mentions.] Are they a long-term solution? Obviously, no. Yet, for the common Joe outside the obvious sphere of Twitter — it is a more simple baby step. In all the words in the paragraph above, the word “gradually” encapsulates the beauty of sachets- bookmarks in Twitter’s case. And that is what Twitter needs — not the magic of depth, but the simple charm of the breadth. Still propelled by the simplicity of constraints — 140 characters in all. To each one, her own ideas. Do you have ideas for Twitter to grow rapidly? Please share them in the comments section. Let us have a conversation. You can “follow” me on Twitter . I share my favorite reading bookmarks and a few more. @KarthRajan — — — — Karthik RajanI enjoy writing at the intersection of analytics and human relationships. there is an ancient, collection of wisdom as couplets in Tamil called each one of the 1330 couplet is exactly 7 words. Was crafted in 3rd century BC. Every generation still quotes from them! Thirukkural.- is how hackers start their afternoons. We’re a part of the family. We are now and happy to opportunities. Hacker Noon @AMI accepting submissions discuss advertising &sponsorship To learn more, , , or simply, read our about page like/message us on Facebook tweet/DM @HackerNoon. If you enjoyed this story, we recommend reading our and . Until next time, don’t take the realities of the world for granted! latest tech stories trending tech stories

Stories, News & Mentions about GIGAOM

Gigaom’s Stories on HackerNoon

Gigaom's News & Mentions