Future of AI-powered automation in business
-
Upload
louis-dorard -
Category
Technology
-
view
1.582 -
download
1
Transcript of Future of AI-powered automation in business
![Page 1: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/1.jpg)
Future of AI-powered automation in business
@louisdorard #APIdays - December 9, 2015
![Page 2: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/2.jpg)
AI is everywhere
![Page 3: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/3.jpg)
![Page 4: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/4.jpg)
![Page 5: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/5.jpg)
![Page 6: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/6.jpg)
![Page 7: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/7.jpg)
![Page 8: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/8.jpg)
![Page 9: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/9.jpg)
![Page 11: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/11.jpg)
![Page 12: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/12.jpg)
![Page 13: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/13.jpg)
![Page 14: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/14.jpg)
Lars Trieloff
@trieloff
(see source)
![Page 15: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/15.jpg)
Amazon for David Jones (@d_jones, see source)
![Page 16: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/16.jpg)
Amazon for David Jones (@d_jones, see source)
![Page 18: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/18.jpg)
How does it work?
![Page 19: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/19.jpg)
Data + Machine Learning
![Page 20: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/20.jpg)
![Page 21: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/21.jpg)
Bedrooms Bathrooms Surface (foot²) Year built Type Price ($)
3 1 860 1950 house 565,0003 1 1012 1951 house2 1.5 968 1976 townhouse 447,0004 1315 1950 house 648,0003 2 1599 1964 house3 2 987 1951 townhouse 790,0001 1 530 2007 condo 122,0004 2 1574 1964 house 835,0004 2001 house 855,0003 2.5 1472 2005 house4 3.5 1714 2005 townhouse2 2 1113 1999 condo1 769 1999 condo 315,000
![Page 22: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/22.jpg)
![Page 23: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/23.jpg)
Bedrooms Bathrooms Surface (foot²) Year built Type Price ($)
3 1 860 1950 house 565,0003 1 1012 1951 house2 1.5 968 1976 townhouse 447,0004 1315 1950 house 648,0003 2 1599 1964 house3 2 987 1951 townhouse 790,0001 1 530 2007 condo 122,0004 2 1574 1964 house 835,0004 2001 house 855,0003 2.5 1472 2005 house4 3.5 1714 2005 townhouse2 2 1113 1999 condo1 769 1999 condo 315,000
![Page 24: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/24.jpg)
ML is a set of AI techniques where “intelligence” is built by referring to
examples
![Page 25: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/25.jpg)
![Page 26: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/26.jpg)
“Weak AI” vs. “Strong AI”
![Page 27: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/27.jpg)
27
Ever yday use cases
• Real-estate
• Spam
• Priority inbox
• Crowd prediction
property price
email spam indicator
email importance indicator
location & context #people
Zillow
Gmail
Gmail
Tranquilien
![Page 28: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/28.jpg)
28
Business use cases
• Reduce churn
• Cross-sell
• Optimize pricing
• Predict demand
customer churn indicator
customer & product purchase indicator
product & price #sales
context demand
RULES
![Page 29: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/29.jpg)
–Katherine Barr, Partner at VC-firm MDV
"Pairing human workers with machine learning and automation
will transform knowledge work and unleash new levels of human
productivity and creativity."
![Page 30: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/30.jpg)
Decisions from predictions
![Page 31: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/31.jpg)
1. Descriptive
2. Predictive
3. Prescriptive
31
Phases of data analysis
![Page 32: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/32.jpg)
1. Show churn rate against time
2. Predict which customers will churn next
3. Suggest what to do about each customer (e.g. propose to switch plan, send promotional offer, etc.)
32
Churn analysis
![Page 33: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/33.jpg)
1. Show returned goods against {type, customer segment}
2. Predict risk shopper will return goods
3. ?
33
E- commerce returns
![Page 34: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/34.jpg)
“Suggest what to do about each customer” → prioritised list of actions, based on…
• Customer representation + context
• Churn prediction & action prediction
• Uncertainty in predictions
• Revenue brought by customer & Cost of actions
• Constraints on frequency of solicitations34
Churn analysis
![Page 35: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/35.jpg)
35
Pric ing optimisat ion
Again, from David Jones (@d_jones, see source)
![Page 36: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/36.jpg)
Decide price given product and context…
• For several price candidates (within constrained range):
• Predict # sales given product, context, price
• Multiply by price to estimate revenue
36
Pric ing optimisat ion
![Page 37: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/37.jpg)
Decide price given product and context…
• For several price candidates (within constrained range):
• Predict 95%-confidence lower bound on # sales given product, context, price
• Multiply by price to estimate revenue
37
Pric ing optimisat ion
![Page 38: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/38.jpg)
1. Show past demand against calendar
2. Predict demand for [product] at [store] in next 2 days
3. Suggest how much to ship
• Trade-off: cost of storage vs risk of lost sales
• Constraints on order size, truck volume, capacity of people putting stuff into shelves
38
Replenishment
![Page 39: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/39.jpg)
• Context
• Predictions
• Uncertainty in predictions
• Constraints
• Costs / benefits
• Competing objectives (⇒ trade-offs to make)
• Business rules39
Decis ions are based on…
![Page 40: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/40.jpg)
40
Who per forms better?
+vs.
Star Wars: The Flat Awakens by Filipe de Carvalho
vs.
![Page 41: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/41.jpg)
41
AI + Human per form better
+
![Page 42: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/42.jpg)
42
Human alone per forms better : dex terit y
![Page 43: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/43.jpg)
43
AI alone per forms better : replenishment
![Page 44: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/44.jpg)
Decisions are faster, cheaper, and better
44
AI alone per forms better : replenishment
Again, from Lars Trieloff @trieloff (see source)
Decision Quality
Status Quo Predictive Prescriptive Automation
Dec
isio
n qu
alit
y
![Page 45: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/45.jpg)
1. Descriptive analysis
2. Predictive analysis
3. Prescriptive analysis
4. Automated decisions
45
B eyond prescr ipt ive analysis
![Page 46: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/46.jpg)
• Spam filter → decide to skip inbox
• Autonomous Vehicles → decide who to kill
46
Autonomous decis ion-mak ing systems
⇒ “Tool AI” vs “High-stakes autonomous AI”
![Page 47: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/47.jpg)
47
Autonomous Vehicles
![Page 48: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/48.jpg)
• Morality in decision-making algorithm:
• Minimize loss of life
• Account for probabilities of survival, age of occupants…→ optimal formula?
• Sacrifice owner?
• “People are in favor of cars that sacrifice the occupant to save other lives—as long they don’t have to drive one themselves.”
48
Autonomous Vehicles
![Page 49: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/49.jpg)
• Need wide acceptation to get adoption and provide benefit (e.g. save lives with AVs)
• “The public is much more likely to go along with a scenario that aligns with their own views”
• What will the public tolerate? → experimental ethics
• Similar issues whenever AI decides for us and impacts many
⇒ “Domain-specific/business rules” in decision making49
H igh-stakes autonomous AIs
![Page 50: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/50.jpg)
Role of APIs
![Page 51: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/51.jpg)
51
Communication bet ween AIs
01000101101
![Page 52: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/52.jpg)
Software components for automated decisions:
• Create training dataset from historical data (merge sources, aggregate…)
• Provide predictive model from given training set (i.e. learn)
• Provide prediction against model for given context
• Provide optimal decision from given contextual data, predictions, uncertainties, constraints, objectives, costs
• Apply given decision52
S eparation of concerns
![Page 53: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/53.jpg)
Software components for automated decisions:
• Create training dataset from historical data (merge sources, aggregate…)
• Provide predictive model from given training set (i.e. learn)
• Provide prediction against model for given context
• Provide optimal decision from given contextual data, predictions, uncertainties, constraints, objectives, costs
• Apply given decision53
Operations Research component
![Page 54: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/54.jpg)
Software components for automated decisions:
• Create training dataset from historical data (merge sources, aggregate…)
• Provide predictive model from given training set (i.e. learn)
• Provide prediction against model for given context
• Provide optimal decision from given contextual data, predictions, uncertainties, constraints, objectives, costs
• Apply given decision54
M achine Learning components
![Page 55: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/55.jpg)
Software components for automated decisions:
• Create training dataset from historical data (merge sources, aggregate…)
• Provide predictive model from given training set (i.e. learn)
• Provide prediction against model for given context
• Provide optimal decision from given contextual data, predictions, uncertainties, constraints, objectives, costs
• Apply given decision55
Predic t ive APIs
![Page 56: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/56.jpg)
56
Predic t ive APIs
![Page 57: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/57.jpg)
The two phases of machine learning:
• TRAIN a model
• PREDICT with a model
57
Predic t ive APIs
![Page 58: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/58.jpg)
The two methods of predictive APIs:
• TRAIN a model
• PREDICT with a model
58
Predic t ive APIs
![Page 59: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/59.jpg)
The two methods of predictive APIs:
• model = create_model(‘training.csv’)
• predicted_output = create_prediction(model, new_input)
59
Predic t ive APIs
![Page 60: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/60.jpg)
Amazon ML
BigML
Google Prediction
PredicSis
…
60
Providers of REST http Predic t ive APIs
![Page 61: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/61.jpg)
Going further
![Page 62: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/62.jpg)
• Define desired and acceptable behaviour→ objectives and constraints/bounds
• Monitor accuracy & bottomline
• Self-monitoring & anomaly detection→ thresholds and fallbacks
62
Ensuring per formance of autonomous AI systems
![Page 63: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/63.jpg)
63
Per formance guarantees?
“construction worker in orange safety vest is working on road”
95%-accurate scene description
![Page 64: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/64.jpg)
64
Per formance guarantees
“black and white dog jumps over bar”
95%-accurate scene description
![Page 65: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/65.jpg)
65
Per formance guarantees
“a young boy is holding a baseball bat”
95%-accurate scene description
![Page 66: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/66.jpg)
66
Per formance guarantees
“a young boy is holding a baseball bat”weapon
SIR, DROP THE WEAPON!
![Page 67: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/67.jpg)
• Lars Trieloff: “Business reasons for automating decisions”
• Daniel Kahneman: “Thinking, Fast and Slow”
• Tom Dietterich: “Artificial Intelligence Progress”
• MIT Technology Review: “Why Self-Driving Cars Must Be Programmed to Kill”
• Conference: PAPIs Connect67
Learn more
![Page 68: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/68.jpg)
![Page 70: Future of AI-powered automation in business](https://reader034.fdocuments.us/reader034/viewer/2022050900/588409571a28ab884b8b7167/html5/thumbnails/70.jpg)
• Free ML resources: louisdorard.com
• PAPIs updates: @papisdotio