Feb 20, 2020

Ai Business

layout: post title: AI Business Approach


Artificial Intelligence - Intelligence is the benefit for your Client and what you want to sell. Artifical is the challenge, as all models are wrong but sometimes useful.

  • Business Plan / Canvas
    • Either an identified Problem
      • Instrumentation / Control
      • Repetition / Automation / Acceleration
      • Communication / Dialogue
      • Recognition
    • Or identified Technology
      • In search of a problem
      • In search of a market
      • Making sense of a morph dataset
      • Technology becomes a constraint in the solution space
    • Customer
      • Value
      • Control
      • Personalization
      • Closer to cognitive dialogue
    • Cost Case
      • Availability of quality Data, defined outcome, cost and risk
      • Constraints definition
      • Frequent re-train of images, testing and distribution of updates
      • Team and People dependencies
        • Better organize Knowledge in the System and structured Data
        • Churn of People
      • Support(see Testing), SmartServices Potential and Necessity
      • Risk Management
    • Organization
      • Implications for the Process Layer
        • Internal automation
        • New labour
      • Responsive and Agile
      • IT Utilities and strategic Assets
    • Price
      • Standard or Project
      • AI implies disruptive flexible invests and payment scheme
      • It’s about automation and control, less discounts
      • No intelligence without automation, price for appropriate infrastructure
  • Data
    • Sourcing of quality Data
    • Control of instrumentation, sampling entities and diversity
    • Ownership of Data and Partaking agreements
    • Understanding of synthetic Data, modifications applied and state of sources
    • Graph or semantic structure allow for alternative and re-use
    • A morph data with unsupervised - What does it tell you?
    • Capability to conduct updates of Datasets
  • Foundation for Algorithms
    • Next Generation Capabilities
      • Visual Systems
      • Operational pipes
      • Generative Models
    • Math, Engineering, Stats, Distributions
      • Owned key developments
      • Derive criteria for an industry platform
      • Modeling (Including Virtual Worlds for Testing)
    • Capacities for Problem Solving and Reasoning
      • Architecture (Central, Cloud, Edge, IoT…)
    • Updates, Release Management, DevOps
  • Testing
    • Really understand the function intended
    • Behavior (AI typically is a black box)
    • Stress cases and fulfillment of QoS
    • Reproduction of behavior including Legal Analysis in Virtual World after incidents
  • Dialogue
    • Machine Man
    • Machine Machine, Signaling, APIs
    • Machine Machine Dialogue including Man, understanding yields trust