Data and AI: The largest technology disruption in 240 years - Sam Lightstone CTO for Data & IBM Fellow - Sam Lightstone(2)
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
Data and AI: The largest technology disruption in 240 years Sam Lightstone CTO for Data & IBM Fellow #ThinkLisboa
Disclaimer IBM’s statements regarding its plans, directions, and intent are subject to change or withdrawal without notice at IBM’s sole discretion. Information regarding potential future products is intended to outline our general product direction and it should not be relied on in making a purchasing decision. The information mentioned regarding potential future products is not a commitment, promise, or legal obligation to deliver any material, code or functionality. Information about potential future products may not be incorporated into any contract. The development, release, and timing of any future features or functionality described for our products remains at our sole discretion. Information presented and discussed during this meeting may be both IBM and client confidential. The agreements signed by members of the Technical Advisory Board govern usage of any and all information discussed and shared.
The world’s largest personal transport company owns NO vehicles.
The world’s largest accommodations provider 12 owns no real estate.
Photo from coursera.com Coursera over 30 million users no classrooms 13
The biggest disruption in 240 years IBM Analytics
The most disruptive innovations of the past 600 years IBM Analytics
AI is creating the largest mass automation since the advent of the steam engine in 1776. IBM Analytics
Why now? IBM Analytics
IBM Analytics
How does machine learning relate to data science and AI? Artificial Intelligence Linear regression Logistic regression Machine Linear Discriminant Analysis Learning Classification and Regression Trees Naive Bayes Deep K-Nearest Neighbors Learning Learning Vector Quantization Support Vector Machines Data Bagging and Random Forest Science Gradient Boosting Data access, & Artificial Neural Networks (many kinds) preparation 19
The era of Machine Learning IBM Analytics
“Our relationship to computers has changed. Instead of programming them we now show them and they figure it out.” Geoffrey Hinton “The godfather of deep learning” See https://youtu.be/-eyhCTvrEtE. 34m.50s IBM Analytics
AI is getting real You want to debate that? IBM Analytics
24
Philyra system Daub hired IBM to create AI for perfume. Produces new inventive perfumes Consumes large amounts of information about the formulas of existing fragrances, consumer data, regulatory information, etc. 25
The AI Ladder Multi-Cloud Data Architecture INFUSE – Automate and scale across your processes TRUST– Archive trust and transparency in outcomes ANALYZE – Scale insights with ML everywhere ORGANIZE– Create a trusted analytics foundation COLLECT – Make data simple and accessible Data of every type: your critical data assets no matter where they physically are Think 2019 / DOC ID / February 12, 2019 / © 2019 IBM Corporation
IBM Data & AI Portfolio Everything you need for Enterprise AI, on any cloud Pre-built Use Cases Watson Applications The Ladder to AI Build Run Manage Watson Watson Watson Machine Studio OpenScale Learning Hybrid Data Management Data Governance & Integration Open source meets a multicloud, working as ONE Db2 Family InfoSphere Family Multicloud Data & AI Platform IBM Cloud Pak for Data 27
Watson Studio Pre-Integrated Tools, Algorithms, Libraries for Data Science, ML/DL Tools • Best of breed open source & IBM tools • Code (R, Python or Scala) and no-code/visual Decision modeling tools Optimization • Most popular open source frameworks Machine Learning Runtimes Deep Learning Runtimes • IBM best-in-class frameworks • Container-based resource management Scalable Infrastructure • On IBM Cloud: Elastic pay as you go CPU/GPU use Model Lifecycle Management Think 2019 / 6974 / February 14, 2019 / © 2019 IBM Corporation 28 28
Use Cases for Machine/Deep Learning Cyber Defense Fraud Detection Drug Discovery Loan Face Chatbots IoT Approval Medical Recognition Decision-Making Recommender Weather systems Disease Forecasting Diagnostics Targetted marketing Robotics Advanced Physics Research Supply Chain Sales Media Analytics Management Forecast 29
Fun fact: ~75% of practical ML projects for business operate on text and structured data (i.e. relational or JSON). Small amount on audio. Very small on image & video. Estimate, as of March 2019 IBM Analytics
What if you could tap into all of your critical data assets no matter where they physically are? Think 2019 / DOC ID / February 12, 2019 / © 2019 IBM Corporation
IBM Data Virtualization Unified data asset catalog, lineage Unified access control and and provenance security policies Data Virtualization [+ caching layer] Data Warehouses & Big Data Marts (Hadoop) Spreadsheets & Relational Text files Databases No SQL Locations: Private and public clouds, standalone systems, worldwide. Think 2019 / DOC ID / February 12, 2019 / © 2019 IBM Corporation
IBM Data Virtualization Rich application capabilities • Connect to Data Virtualization with your favorite SQL apps and tools • RStudio, Jupyter Notebook, Cognos, Tableau, Microstrategy • Db2 SQL and driver compatible IBM Analytics
Demo IBM Analytics
Meet Max —Name: Max von Datacrunch —Position: CIO for Capital Mercantile Bank —Data: 12 Petabytes under management —Pet peeve: Spends 600M USD on IT, but can’t see a global view of his data. —Pastimes: none —Medical: • Age 42 (looks 64). • Blood pressure 180/120 • LDL cholesterol: 210 mg/dl Demo Think 2019 / DOC ID / February 12, 2019 / © 2019 IBM Corporation
The Capital Mercantile Bank * What you’ll see — Data Virtualization over widely distributed data — SQL editor inside ICP for Data — Data Science Experience with R Studio query over distributed databases — Plot.ly data visualization The setup (actual!) — Scale: 8 distributed data sources — Format: MySQL, Informix, Db2, Oracle and others. — Global locations: Hursley UK, Toronto CAN, San Jose USA — Hardware: Mix of Intel and ARM CPUs. The transaction data — 2 ½ Years worth of data — 1,500 data point per database per day The Capital Mercantile Bank Talented. Trusted. Global. Demo Think 2019 / DOC ID / February 12, 2019 / © 2019 IBM Corporation * fictional bank
We are the technology company for the enterprise IBM Analytics
Thank You Sam Lightstone 39
About the Speaker Sam Lightstone is IBM Chief Technology Office (CTO) for Data, an IBM Fellow and Master Inventor. He loves to help customers solve real problems and help IBM invent the future. He has over 60 patents and is widely published. In his spare time he is a an avid guitar player and fencer. @samlightstone
You can also read