AI Flashcards
What is the AAAI definition of AI
An understanding of the mechanisms underlying thought and their embodiment in machines
What are the possible AI frontiers?
Increases mental automation Next gen computing Cognitive computing Vision recognition Language translation Autonomous vehicles
Artificial general intelligence
Automation or augmentation
Tell me about VA’s
Clearly where organisations see AI
May transform retail and communication- can buy from Alexa
Apple may launch a home device - increased connectivity and machine learning
Echo look- contentious
Challenge of big data
Look at quantities of data beyond human comprehension and traditional computing
Volume- can look into great depth of data
Velocity- can pre filter
Variety- analyse all types from structured to unstructured- cloud AI
What is IBM Watson
1 billion business unit dedicated to the development and commercialisation of cloud delivered cognitive innovations
Machine learning- observe interpret and analyse
Why is Watson struggling to create revenue?
Reducing AI to a commodity service has proven difficult
Now other players in the field
Repeating cycles of techno economic revolution
What did McKinsey say about data driven medicine
Could save 300 billion using AI
What is enlitic
Uses deep learning to provide start of the art clynical support
Screening- can screen 50% more accurately for lung cancer than radiologists
Interpret a medical image 10,000 times faster - link to trans humanism we need to step up
What is Watson for oncology
Provides personalised cancer treatments
What are the landmarks for autonomous vehicles
Expected semi autonomous vehicles by 2020
By 2025 high autonomy vehicles likely
By 2030 fully autonomous may exist
What are the 3 eras of automation davenprot and Kirby
1) take away dangerous
2 take away full
3 take away decisions
What does grace 2017 future of humanity institute find
AI will outperform humans in the next 10 years
Translate by 2024
Drive trucks by 2027
Capable of being a surgeon 2053
50% chance AI will be better than humans in more or less everything in 45 years
In 2015 it was estimated to take 12 years for AI to beat a human at go but google have done this with deep mind in just 2
IBM claim
Chat bots will help businesses save 8 billion per year b6”y 2022
4 minutes time saving
70 cents saved per chat or interaction
93% chat not success rate in banking
Why has googles deep mind had problems with the NHS?
Received 1.6 million medical records
Reviewing data records that weren’t on kidney disease which was its intended purpose
