An intro into AI and how business leaders should use it
1 like271 views
The document discusses the evolution and application of artificial intelligence (AI) in business, emphasizing that AI is a tool rather than a standalone business model. It outlines key rules for integrating AI, such as the necessity of having the right and protected data, and the importance of a holistic approach involving company structure and culture. Additionally, it highlights the role of regulation and ethical considerations in AI development and deployment.
![LutzFinger.comLutzFinger.com
LutzFinger
Regulation Considerations
Ability to Explanation
“[...] the controller shall provide [...]
the following information: the
existence of the automated
decision-making and [...] meaningful
information about the logic involved
and the envisaged consequences of
such processing for the data
subject.”
Art. 13-15 & 22 Regulation (EU)
2016/679](https://image.slidesharecdn.com/barcelonadata-sep20192-190927164719/85/An-intro-into-AI-and-how-business-leaders-should-use-it-39-320.jpg)
An intro into AI and how business leaders should use it
- 3. LutzFinger.comLutzFinger.com LutzFinger Is Angela Merkel A Boy? Campaign Image 2017
- 6. LutzFinger.comLutzFinger.com LutzFinger AI started in 1956 Source: Wikipedia https://www.futuretimeline.net/forum/topic/21039-broken-promises-empty-threats-the-evolution-of-ai-in-the-usa-1956-1996/
- 7. LutzFinger.comLutzFinger.com LutzFinger 1st. Wave: ML Source: Wikipedia BEFORE: ● Program Each Step NOW: ● Symbolic Artificial Intelligence > Rule Based Learn how to Code with Swift
- 8. LutzFinger.comLutzFinger.com LutzFinger 2nd. Wave: More Rules Source: Wikipedia BEFORE: ● AI limited to small applications like Checkers NOW: ● Aiming to build systems to augment the knowledge of doctors “We know more than we can tell” Michael Polany 1958
- 9. LutzFinger.comLutzFinger.com LutzFinger 3rd. Wave: Computer Designs Rules Find likelihood for a SERIOUS DELINQUENCY (2y past due for mortgage) AI Optimizes to find the best Decision. AI writes the if-then Rules.
- 12. LutzFinger.comLutzFinger.com LutzFinger RULE 1: AI is a Tool Not A Business Model
- 13. LutzFinger.comLutzFinger.com LutzFinger AI Can Create New Products https://eng.uber.com/atg-dataviz/ https://www.symrise.com/newsroom/article/breaking-new-fragra nce-ground-with-artificial-intelligence-ai-ibm-research-and-symr ise-are-workin/ New Products Can Create New Business
- 14. LutzFinger.comLutzFinger.com LutzFinger Success is based on your Business Model
- 15. LutzFinger.comLutzFinger.com LutzFinger Digitalization Has Changed Business Levers From: https://stratechery.com/2015/aggregation-theory/ Example: Media Industry Media Publication 1 Media Publication 2 Media Publication 3 Many Suppliers Few Suppliers
- 16. LutzFinger.comLutzFinger.com LutzFinger The Levers • Direct Relationship with Users • They Own the Relationship • No One is between Demand and them • Zero Marginal Costs For Serving Incremental Users • No COGS • No Distribution Cost • No Transaction Cost • Abundant Supply • With either zero marginal acquisition cost • OR acquisition cost that scales Those are attractive terms for the business. Data improves over time the service and creates a moat for competition. But data does not create the business model. Further reading: https://stratechery.com/
- 17. LutzFinger.comLutzFinger.com LutzFinger Business Imperative • Amazon One Vendor • Tesla Over The Air Software Update • Netflix Originals • Disney New Streaming Service • Uber city-per-city investment in Supply • Facebook Instant Articles Who Is Pushing Into Your Space?
- 18. LutzFinger.comLutzFinger.com LutzFinger Tools Create New Product & New Business Tools Help To Optimize Your Core Biz. https://eng.uber.com/atg-dataviz/ https://www.symrise.com/newsroom/article/breaking-new-fragra nce-ground-with-artificial-intelligence-ai-ibm-research-and-symr ise-are-workin/ ● Banking: Card Fraud Detection ● Banking: Credit Scoring ● Media: Content Recommendation ● Health Care: Fraud Detection ● Medicine: Image Processing ● Medicine: Outliers Detection ● Retail: Likelihood to Buy ● Books: Marketing Planning ● Manufacturing: Failure Prediction ● Manufacturing: Optimization ● Insurance: Likelihood & Pricing ● Transportation: Route Planning ● Energy: Grid Utilization
- 19. LutzFinger.comLutzFinger.com LutzFinger Insurance ● Damage Estimations: Using Deep Learning & Satellite Images to understand damage of the Camp Fire in 2018 ● Insurance Premiums: Praedicat mines literature to extract metadata about bodily harm that can be attributed to certain materials. https://www.space.com/41403-california-wildfires-2018-photos -from-space.html ● Claim Management & Indemnity Prediction: Digital workflows improve workflows and predictions (AI Insurance)
- 20. LutzFinger.comLutzFinger.com LutzFinger Oil & Gas ● Clean up data from pressure transient analysis (PTA) using Singular Spectrum Analysis (SSA) and a supervised learning ● Methode can replace the manual task of identifying the flow regime. https://doi.org/10.1145/3292500.3330661 https://www.kdd.org/kdd2019/accepted-papers/view/structured-noise-detection-application-on-well-test-pressure-derivative-dat
- 21. LutzFinger.comLutzFinger.com LutzFinger Retail Credit Loan Applications (RNN over debit or a credit card transactions) 1) More accurate than traditional 2) Less human overhead (raw transaction data >> no feature needed) 3) Fast as no additional customer input is needed (transactional data) 4) Hard to forge signals (transactional data) 5) Works for customers without credit history https://doi.org/10.1145/3292500.3330693
- 22. LutzFinger.comLutzFinger.com LutzFinger Customer Care 3 Disclaimer: I am an angel investor in ultimate and i2x. https://medium.com/sap-io/how-an-ai-powered-contact-center-solution-ca n-save-62-billion-in-lost-sales-b53e3c906cb6
- 23. LutzFinger.comLutzFinger.com LutzFinger Fitbit for Cows Helping to manage health from cows to improve milk production. The only difference between cows that produce 30 liters of milk a day and those that produce 10 liters was the animal’s health. https://itunes.apple.com/app/ida-by-connecterra/id1246795438 https://www.connecterra.io/ida/registration-for-ida-event-at-neu-hope-dairy-usa/ida-dashboard-en/
- 24. LutzFinger.comLutzFinger.com LutzFinger RULE 2: Not More Data But Right & Protected Data
- 25. LutzFinger.comLutzFinger.com LutzFinger AI needs Data from Marketing Material of Ursa Space Systems RSMetrics https://medium.com/@RSMetrics/metalsignals-tracking-the-worlds-shadow-sup ply-with-satellite-imagery-f4f1c67717c8
- 26. LutzFinger.comLutzFinger.com LutzFinger You Need to Protect Your Data
- 27. LutzFinger.comLutzFinger.com LutzFinger Data Can Be Shared - Central
- 28. LutzFinger.comLutzFinger.com LutzFinger … and Decentral > Federated Learning APPLIED FEDERATED LEARNING: arXiv:1812.02903v1 Step 1: Train Model in Cloud Initial Data Model
- 29. LutzFinger.comLutzFinger.com LutzFinger Federated Learning APPLIED FEDERATED LEARNING: arXiv:1812.02903v1 Step 1: Train Model in Cloud Step 2: Distribute the Model & Improve Model Local Company X Company Y Company Z Initial Data Model
- 30. LutzFinger.comLutzFinger.com LutzFinger Federated Learning APPLIED FEDERATED LEARNING: arXiv:1812.02903v1 Step 1: Train Model in Cloud Company X Company Y Company Y Step 3: Upload Model not the Data & Merge Models Initial Data Model Step 2: Distribute the Model & Improve Model Local
- 31. LutzFinger.comLutzFinger.com LutzFinger Federated Learning APPLIED FEDERATED LEARNING: arXiv:1812.02903v1 Step 1: Train Model in Cloud Company X Company Y Company Y Step 3: Upload Model not the Data & Merge Models Step 2: Distribute the Model & Improve Model Local Repeat
- 32. LutzFinger.comLutzFinger.com LutzFinger Data has also Diminishing Returns by Claudia Perlich et. al., “Tree Induction vs. Logistic Regression: A Learning-Curve Analysis”, Journal of Machine Learning Research 4 (2003) 211-255
- 33. LutzFinger.comLutzFinger.com LutzFinger There is a Way to Jumpstart Your Data Needs
- 34. LutzFinger.comLutzFinger.com LutzFinger Transfer Learning Transfer Learning is a strategy where a model used for one tasks (e.i. recognition of animals) used as starting point for the second task (e.i. recognition of faces) Step 1: Peel Back Layer Step 2: Add Layer & Re-Train with Your Data
- 35. LutzFinger.comLutzFinger.com LutzFinger Filter Layers Your Model Knows Animals See: https://research.googleblog.com/2015/06/inceptionism-going-deeper-into-neural.html?m=1
- 36. LutzFinger.comLutzFinger.com LutzFinger Transfer Learning at Work Combining historical mining and exploration data with NASA data and use pre-trained deep learning networks to identify several target zones with the highest potential for gold mineralization in Nevada’s Jerritt Canyon district.
- 37. LutzFinger.comLutzFinger.com LutzFinger RULE #3: Regulation & Ethics Concerns Should Guide You Not Stop You
- 38. LutzFinger.comLutzFinger.com LutzFinger The Latecomer Argument “AI is the most serious threat to the survival of the human race” Elon Musk
- 39. LutzFinger.comLutzFinger.com LutzFinger Regulation Considerations Ability to Explanation “[...] the controller shall provide [...] the following information: the existence of the automated decision-making and [...] meaningful information about the logic involved and the envisaged consequences of such processing for the data subject.” Art. 13-15 & 22 Regulation (EU) 2016/679
- 40. LutzFinger.comLutzFinger.com LutzFinger One solutions: LIME ( local interpretable model-agnostic explanations) https://github.com/marcotcr/lime AND https://homes.cs.washington.edu/~marcotcr/blog/lime/ Predict on PermutationTake Data Point(s) = Local Permute Data Points (+ Noise) Use distance and other similarity metrics to understand the most important (not permuted) features.
- 41. LutzFinger.comLutzFinger.com LutzFinger Biased Data Concerns Trainings Data Label “dogs” Label “cats” Predictive Model Input Data Class “cats”
- 44. LutzFinger.comLutzFinger.com LutzFinger AI needs the whole company ● Business Model ● Data ● People ● Infrastructure ● Organizational Set Up And… Leadership
- 45. LutzFinger.comLutzFinger.com LutzFinger Example: China People Leadership Data Infrastructure doi: 10.1038/d41586-019-02360-7 AND https://flia.org/notice-state-council-issuing-new-generation-artificial-intelligence-development-plan/ https://macropolo.org/chinas-ai-talent-base-is-growing-and-then-leaving/ https://www.wired.com/story/china-catching-up-us-in-ai-research/
- 48. LutzFinger.comLutzFinger.com LutzFinger Your Cookbook 1. AI is a Tool Not A Business Model a. Not everything is a unicorn b. Biz improvements are a good start 2. Not More Data But “Right” & “Protected” Data a. Own / protect the access b. Use federated learning to share models c. There are diminishing returns in data 3. Regulation Should Guide You Not Stop You a. Use LIME to understand the model b. Test for Biases 4. Think Holistically a. Hire data scientists and engineers b. Set a AI minded culture
- 50. LutzFinger.comLutzFinger.com LutzFinger Lutz Finger Disclaimer This content is solemnly my own opinion and not necessarily the opinion of any of my employer or former employers.