Artificial stupidity
Codemotion Roma 2019-03-23
Your Host Today
What We Will Cover
Coming of Age
Case Studies
Analysis
Conclusions
Why Me?
20+ years in the industry
Interested in the area since the 90s
Read a bunch of blog posts
Extremely muscular body
In Memoriam Sergio Leone
Sad HIll Unearthed
Early AI
Pioneers
Quite Optimistic
In the literature (US)
In the Literature (EU)
Coming of Age
Chess
Extra, Extra! Humans Defeated!
Automated Theorem Proving
In the Literature (us)
In the Literature (Japan)
Case Studies
1: Robots
Industrial Robots
space Robots
BattleBots
Don't give them the kill switch!
Actual robots
Not so Intelligent
Not real bots
Not so versatile after all
Exo drives
Humans are great value after all
Why the emphasis on robots?
The robot myth has been used for centuries to discipline the workforce.
Workers might be replaced first by a steam machine, next by an industrial machine and now by an intelligent machine.
-- Antonio Casilli, 2019
2: Automatic Programming
Who reports bugs?
Not a new thing
3: self-Driving car
Improvements are awesome
state of the art sensors
vs primate in metal box
Ape wins
Self-Driving Car Manufacturers
Visionaires?
The automobile of the day-after-tomorrow will not be driven by the owner, but by itself.
It may be one day a serious offence to drive an automobile on a public highway
4: Deep learning
Using Two Neurons (2004)
Alphazero
The game of chess represented the pinnacle of AI research over several decades
Alphazero on your laptop?
Not so astonishing?
The End of the Line?
I think we need to consider the hard challenges of AI and not be satisfied with short-term, incremental advances.
Realistically, deep learning is only part of the larger challenge of building intelligent machines.
Maximization
The power of a neuron
5: Face Recognition
Pattern Recognition too!
Scrambled Images
"Live" Demo
One pixel can fool a nn
One pixel attack for fooling deep neural networks
A tragedy in four acts
Jaywalking?
Easy to Misuse
Too Many Giraffes?
Not easy to escape?
London: Police fine man £90 for hiding from face recognition
Analysis
Optimization Problems
AIs are great at optimizing multiple parameters in many-d
Everything becomes an optimization problem
Ethics are important
Politics as an Optimization problem
Maximize votes
Minimize inequality?
Maximize power?
Minimize dissent?
Philosophy as an optimization problem
Maximize coherence?
Minimize first principles?
What should we ask?
Religion as an optimization problem?
...
Don't even know where to start
Maximization of unfairness
AI as tools
Employment
AI uprising
Conclusions
AI is awesome, but not that awesome
Using AIs is not a blank check
Think about what you are doing
Don't give AIs the kill switch
