O slideshow foi denunciado.
Utilizamos seu perfil e dados de atividades no LinkedIn para personalizar e exibir anúncios mais relevantes. Altere suas preferências de anúncios quando desejar.

Responsible AI

Speaker: Amy Hodler, Graph Analytics and AI Program Manager, Neo4j

  • Seja o primeiro a comentar

Responsible AI

  1. 1. Responsible AI Amy Hodler Analytics and AI Program Manager Neo4j
  2. 2. AI Gone Mean 2 Microsoft Tay Twitter Bot Learned from users how to respond Learned offensive language and slurs Treated loud, outrageous opinions as the norm Source: The Verge, 2016
  3. 3. Biased AI 3 Recruiting Tools Amazon recruiting tool shut down for bias against women after it codified discriminatory practices due to narrow data sets Sources: Thompson Reuters, 2018, The Verge 2019, The Seattle Times, 2019, Gender Shades Project 99-100% 65-79%93-98% 88-94% Recognition AI Calls for regulation on use of facial recognition after consistently higher error rates for darker-skinned and female faces
  4. 4. Inappropriate AI? 4 China Social Credit System Ranks citizens’ behavior to determine their social and credit worthiness 1.4B people will have a score by 2020 which will impact their social and economic rights Source: CBS News, 2018
  5. 5. As creators of artificial intelligence systems, we have a duty to guide the development and application of AI in ways that fit our social values Responsible AI Accountability Fairness Public Trust
  6. 6. AI Needs Context
  7. 7. Artificial Intelligence is the WHAT Computer processes that have learned to accomplish specific tasks in ways that mimic human decisions PROBABILISTIC Algorithms train models via specific examples and progressive improvements without explicit direction EATS LOTS OF DATA 7 Machine Learning is the HOW
  8. 8. Decisions Require Context and Connections 8 We observe, collect adjacent data, and make connections We process the connections and to learn and make informed, in-context decisions We make tens of thousands of decisions daily, most of which depend on surrounding circumstances and context. 45
  9. 9. AI Requires Context and Connections, Too 9 45 AI must access and process a great deal of contextual, connected information • Learn from adjacent information • Make and refine judgements • Adjust to circumstances The fastest, most reliable way to manage data connections is with graph technology
  10. 10. Graph Is Accelerating AI Innovation 10 4,000 3,000 2,000 1,000 0 2010 2011 2012 2013 2014 2015 2016 2017 2018 AI Publications with Graph in Title graph neural network graph convolutional graph embedding graph learning graph attention graph kernel graph completion AI Research Papers Featuring Graph are on the Rise Source: Dimension Knowledge System with over 100 research organizations
  11. 11. 11 AI is Limited Without Context ? Narrowly focused Subpar predictions Limited transparency
  12. 12. Graphs as Context
  13. 13. There Is No Isolated Data in Nature 13 Graphs are built for relationships – with relationships Imbue individual entities with connections as a fabric Enriches data so it is more useful
  14. 14. 14
  15. 15. Neo4j and the Property Graph Model 15 EMPLOYEE name: Amy Peters date_of_birth: 1984-03-01 employee_ID: 7875 COMPANY CITY :HAS_CEO start_date: 2008-01-02 :LOCATED_IN start_date: 2008-01-02 Neo4j invented the property graph in 2002 using a napkin sketch – the connected-data model that still works today. Neo4j built a graph database that can process millions of data connections per second
  16. 16. Graph Context for Responsible AI 16
  17. 17. Situational flexibility Predictive accuracy Fairness Reliability and Explainability Graphs Facilitate Responsible AI 17 Robustness Trustworthiness Incorporating context and connections improves the quality and value of AI systems
  18. 18. $72.5 Billion Opioid Insurance Fraud per Year In frauds rings drugs are improperly prescribed by doctors and filled by cooperating pharmacists, all of whom pocket illegal payments Prescriptions for Peril 18
  19. 19. 19 Graph algorithms reveal clusters of interactions in large networks to detect communities for ML Graph Analysis for Detecting Fraud, Waste, and Abuse in Health Care Data Predicting fraud accurately requires extreme insight into the relationships among entities Prescription Fraud Detection with Graphs and ML
  20. 20. Driverless Cars Must Be Foolproof Tesla autonomous car tricked into changing lanes with stickers It’s disturbingly easy to trick AI into doing something deadly 20
  21. 21. 21 Autonomous Decisions Adjacent data helps widen and deepen the scope of AI systems so they are more broadly applicable in their environments Situational awareness is crucial when context-based learning and actions are part of AI systems 4 5
  22. 22. 22 Gaming the System High-stakes criminals misrepresented and manipulated input data to fly under the radar Detecting evolutionary financial statement fraud
  23. 23. 23 Prevent Data Manipulation When data is stored as a graph, it’s easy to track how it changes, who changes it and where it is used For AI solutions to be viewed as reliable the underlying data needs to be reliable Example from Neo4j Risk Mgmt. Solutions
  24. 24. Past and Current Data Amplifies Bias Data skewed by discrimination and demographics creeps into policing, programs and sentencing To Predict and Serve? Predictive Policing Systems Machine Bias report on COMPAS Software by ProPublica 24 COMPAS Scores at Booking
  25. 25. 25 Reveal and Eliminate Bias Understanding our data can reveal bias inherit in the information, in how it’s collected or in how it’s used to train our models Graphs adds contextual information to our ML data and reveals relationships within data – which are often better outcome predictors than raw data Connected by James Fowler “…data without context is just organized information.” Albert Einstein
  26. 26. 26 Human Interaction is Crucial Boeing fails to incorporate pilot reactions into 737 Max auto-pilot system Too many human errors brought down the Boeing 737 Max
  27. 27. 27 Human Centric AI systems can be over-fitted to tight scenarios and idealized situations that don’t account for the range of human interactions Graphs encapsulate the way we think about the world, making it easier to incorporate human responses and explain outcomes / processes
  28. 28. What’s Next 28
  29. 29. Graphs Already Bring Context to Data Science, Machine Learning and AI 29 Financial Crimes Recommendations Cybersecurity Predictive Maintenance Customer Segmentation Churn Prediction Search & MDM Drug Discovery
  30. 30. Context for AI Will Be Standard 30 The inclusion and use of adjacent information as context for AI will become a standard This will drive more reliable, accurate and flexible AI solutions
  31. 31. “The idea is that graph networks are bigger than any one machine-learning approach. Graphs bring an ability to generalize about structure that the individual neural nets don't have.” Next Major Advancement in AI: Graph Native Learning
  32. 32. Next Major Advancement in AI: Graph Native Learning 32 Implements machine learning in a graph environment Native graph learning will move today’s AI from a rigid, black box approach to extremely flexible, accurate and transparent models Lets users input connected data Learns while preserving transient states Produces outcomes in graph format Enables experts to track and validate AI decision paths More accurate with less data, learning important features
  33. 33. 33 “Coders are the most empowered laborers that have ever existed.” Anil Dash @anildash Glitch CEO Ethical technology activist
  34. 34. 34 Getting Started Training/Modeling Outcomes Add Context to AI Decisions (Knowledge graphs) Stop, Think and Acknowledge Tools for More Responsible AI Know & Track Data (Graphs for data lineage) De-Bias Data (AI Fairness 360 toolkit) Add Relationships to ML Training (Graph feature engineering Counterfactual search) Model Exchanges (ONNX, MAX) Learn/Ask for Help (Algorithmic Justice League)
  35. 35. Human Values Will Impact AI AI guidelines that promote societal values AI solutions will increase situational appropriateness, tamper-proofing, explainability and transparency Faster adoption of AI solutions as they become more trustworthy Sources: NIST, Univ of Oxford, The Verge
  36. 36. Parting Thoughts About the Future of AI
  37. 37. 37 “A lot of times, the failings are not in AI. They're human failings... …if you’re not thinking about the human problem, then AI isn’t going to solve it for you.” Vivienne Ming Executive Chair & Co-Founder, Socos Labs
  38. 38. 38 “Context-awareness is a core requirement … . . .that it has sufficient perception of the user’s environment, situation, and context to reason properly.” Oliver Brdiczka AI Architect, Adobe Professor, Georgia Tech
  39. 39. 39 “AI is not all about Machine Learning. Context, structure, and reasoning are necessary ingredients, and Knowledge Graphs and Linked Data are key technologies for this.” Wais Bashir Managing Editor, Onyx Advisory
  40. 40. April 20-22, 2020 | New York Connect Your Data. Build The Future. graphconnect.com
  41. 41. Responsible AI Amy Hodler @amyhodler amy.hodler@neo4j.com

×