Tech - gusenov/kb GitHub Wiki

misc.

Extremums

History

Tech race

What's next

Wikipedia

MIT Technology Review

Conferences

//build/

  • 2022
    • Build recap (Satya Nadella 2022 Build Keynote)
      • 1) Developer flow
      • 2) Cloud ubiquity
      • 3) App ubiquity
      • 4) Cloud-native
        • Microservices (build, deliver, scale applications faster)
        • APIs
        • Containers (industry strandard to go from code to cloud-native and to scale from zero to infinity)
          • Azure Container Apps
            • fully managed serverless container service
            • built on Kubernetes
            • deploy containerized apps from Docker containers or source code repos
            • apps can scale in response to HTTP traffic or event triggers
        • Serverless
        • Azure Kubernetes Service
          • spin up managed Kubernetes clusters
          • using the community tools like Helm, Kustomize, and GitHub Actions
        • Azure Confidential computing (eliminates the need to de-identify or anonymize protected information because the data is never visible)
      • 5) Unified data
        • We are moving to a world where every app will be intelligent and adapt in real time.
          • Real-time personalization
          • Offer the right product to the right customer at the right time
        • Analytics will become a critical part of the product expirience
        • Databases, analytics, governance (data and privacy officers)
        • Making data integration, data warehousing, MLOps, and BI (data engineers, data scientists, business analysts)
        • Microsoft Intelligent Data Platform (one intelligent data platform)
        • Cosmos DB, SQL Hyperscale, Synaple Link, Azure ML (real-time learning), Power BI, Microsoft Purview
        • Bringing together all data products into one architectural fabric
      • 6) Models as platforms
        • Azure
          • the most powerful AI supercomputer in the world
          • used it to train
        • Large state-of-the-art AI models becoming powerful platforms themselves
          • Microsoft AI
            • Turing - Rich language undestanding
            • Z-Code - 100 languages translation
            • Florence - Breakthrough visual recognition
          • OpenAI
            • GPT - Human-like language generation
            • DALL-E - Realistic image generation
            • Codex - Advanced code generation
        • Applying them (embed them in your applications)
          • Azure Cognitive Services
            • Vision Service
            • Speech Service
            • Language Service
            • Decision Service
            • OpenAI Service
          • Azure Applied AI Services
            • Cognitive Search
            • Form Recognizer
            • Immersive Reader
            • Bot Service
            • Vide Analyzer
        • Use case
          • Knowledge management
          • Summarization for documents and conversations
          • Generate new marketing content
          • Do transformative things
        • Copilot for everything
          • Azure OpenAI Codex model
          • 3D workflows in Maya: add objects in space, modify them in a scene, change the color and attributes of assets, and group, duplicate, position objects in relation to each other
          • simple natural language commands
          • new realm of human-machine interfaces and no-code development
          • without having to navigate a series of menus
          • deep menu commands down to simple, natural sentences
      • 7) Hybrid AI
        • One thing that will surpass the AI supercomputer, it's the supercomputing capacity that will be distributed across all the devices we carry and use throughout our lives
        • In the last two years, created 50X more GPU compute on Windows compared to the biggest AI supercomputer
        • World where every Windows computer will draw on the combined power of CPUs, GPUs, NPUs and even a new coprocessor Azure compute
        • In this hybrid, cloud-to-edge world you will be able to do
          • large-scale training in the cloud
          • and do inference at the edge, and have the fabric work as one.
        • Build cloud-native applications with these rich AI models where they can offload for local inference wherever NPUs are available
        • You should be able to optimize for the device and the local capability you have
        • Hybrid Loop - cross-platform development pattern for building AI experiences that span the cloud and the edge
          • Allows you to make late-binding runtime decisions on whether to run inference on Azure or the local client
          • Can dynamically shift the load between client and cloud
          • ONNX Runtime, Azure ML, AI toolchain
        • Project Volterra - Windows Dev Kit with an ARM CPU and NPU (and it includes native ARM64 Visual Studio and .NET support)
          • Neural processor with best-in-class AI computing capacity
          • Stackable design (for desktops and rack deployment)
      • 8) Low-code/no-code
        • Digital-first company
        • Power Apps
          • Express design
          • Build a prototype in Figma, upload the design into Power Apps and have a working app in seconds, leaving developers to focus on data and backend logic
        • Power Pages
      • 9) Collaborative apps
        • Making apps more contextual and people-centric
        • Microsoft Graph (underlies Microsoft 365)
        • Combine the Microsoft Graph with Microsoft teams, you combine the data that describes how people work together with the place they work together
        • Visual Studio Live Share
        • Co-watch, co-create, co-edit content together
      • 10) Metaverse
        • Immersive experiences in gaming can help create community and connection
        • HoloLens, VR headsets
        • Transforming everything from telehealth, to remote maintenance, to collaborative design and training
        • Dissolving the walls between digital and physical participation
        • Build your own immersive worlds. Build your metaverse experience on Mesh (platform service in Azure)
        • The digitalization of people, places and things and their interactions
        • Industrial metaverse
          • helping companies optimize their operations, and automate, simulate, and predict every business function and process
          • brings distributed workforce together; remote experts can share their technical knowledge with not only one but multiple sites immediately speeding up troubleshooting
        • Digital Twins
          • With the digital twin, you can understand the issues and fix them remotely. Look into it in a virtual space.
        • Azure Percept (notice failures and breakdowns)
        • In the world of a complete metaverse all processes from development to design and testing can be executed in a virtual environment
        • IoT

nature.com

Gartner

Deloitte

InfoQ.com

TIME

Tech Stacks

Evaluation, metrics, benchmarks

Critique

LOLZ