AI Communities
http://people2vec.org/
Keras is a high-level neural networks library, written in Python and capable of running on top of either TensorFlow or Theano. It was developed with a focus on enabling fast experimentation.
Being able to go from idea to result with the least possible delay is key to doing good research.
https://www.holbertonschool.com/ https://keras.io
Use Cases ?Dialog & Discussion
Teaching & Learning
Mood Swing Management
Enrollment
Enterprise Communication
Staff Management
Re-Identification
Self Management
Dating
Semantic Search
Health Care
Personal Coaching
Depression
e-Commerce
Reading
Car Personal AssistantSocial Robots
Things2Vec
Call-Center
Social Personal AssistantsPress secretary
Everyone.AI : “because Everyone deserves AI”
Mission : Help everyone to embrace AI• Business Competitiveness with AI• Education to promote new AI jobs• Universal Basic Income Think Tank
Action : Speakers, Writers, Activists & Communities • AI Articles, Twitter & Facebook daily
Publications• Deep Learning Classes in San Francisco, Paris• Meetups with Exec teams of Fortune 1000• Government
Impact : so far in 2016• > 145 Millions People impacted• > 1.776.274 Jobs impacted• > 50 WW CEO & Execs wake-up calls• > 1000 Hours of Teaching
https://cnnumerique.fr/conference-numerique-franco-allemande-2016https://cnnumerique.fr/wp-content/uploads/2016/12/CNNum_BJDW_Poster_ENG.pdfhttps://github.com/holbertonschool/deep-learning
AI landscape• Most AI experts are part of established AI players : AAIM + F• By 2020 these players will control 60% of the AI applications • AI drives a $14-33 trillion economic impact
• Gartner’s predicts: • AI bots will power 85% of customer service interactions• Digital assistant will “know” you by 2018 • 20% of business content will come from AIs by 2018
Andrew Tonner – 12/10/16 http://www.fool.com/investing/2016/12/10/9-artificial-intelligence-stats-that-will-blow-you.aspx
AI products
• Bots• Robots • IoT• Autonomous driving• Customer Services• …
The rise of AI was predictable - Human/Machine balance• Automation: Time, Money, Wellbeing• Competitiveness : Growth• Augmentation : Super Human
HOW
WHY
http://futurearchitectureplatform.org/news/28/ai-architecture-intelligence/
AI product – quest for the golden key
Does AI solve the problem?
What AI methods are needed?
Where is the data?
How do we validate the results?
How do we maintain/iterate?
What do you want to achieve ?
AI Product life cycle(s)
SkillsAKA features
Data
Evaluate
USE RELEASE
Design
AI life cycleTraditional life cycle
Develop
Test
Feed
Learn
Voice
ChatbotSMS
Web
Phone
Communication Channels – HM Interactions
Train
Validate
ML
DL
NLP
IP
SA
Gesture
http://www.datasciencecentral.com/profiles/blogs/has-deep-learning-made-traditional-machine-learning-irrelevant
Traditional SW product vs. AI products
Product roadmap
Maintenance
Back up/Roll Back
QA – Bug free
Skills roadmap
Scalability
? (unplug)
Training – Accuracy – False positive
WHAT DID THE MACHINE LEARN?DOES IT WORK?
AI product other interrogationsError in traditional SW products:
Bug Developer Fix
Error in AI product:
LEARNED The Mirror Effect
CHALLENGES: • Identify that there is an error • There is no one fix just a learning process• Determine liabilities with empathy and ethic• We don’t know what we don’t know• Are we ready to learn from AI?
?