Instant Expert - Artificial Intelligence 2nd December 2017 Flashcards
Machines that think - Douglas Heaven
AlphaGo Zero - the first computer program to defeat a world champion at the ancient Chinese game of Go.
The number of possible configurations of the board is more than the number of atoms in the Universe.
Previous versions of AlphaGo initially trained on thousands of human amateur and professional games.
AlphaGo Zero skips this step and learns to play simply by playing games against itself, starting from completely random play.
In doing so, it quickly surpassed human level of play and defeated the previously published champion-defeating version of AlphaGo by 100 games to 0.
It is able to do this by using a novel form of REINFORCEMENT LEARNING, in which AlphaGo Zero becomes its own teacher. The system starts off with a neural network that knows nothing about the game of Go. It then plays games against itself, by combining this neural network with a powerful search algorithm. As it plays, the neural network is tuned and updated to predict moves, as well as the eventual winner of the games.
This updated neural network is then recombined with the search algorithm to create a new, stronger version of AlphaGo Zero, and the process begins again.
In each iteration, the performance of the system improves by a small amount, and the quality of the self-play games increases, leading to more and more accurate neural networks and ever stronger versions of AlphaGo Zero.
This technique is more powerful than previous versions of AlphaGo because it is no longer constrained by the limits of human knowledge. Instead, it is able to learn tabula rasa (blank slate) from the strongest player in the world: AlphaGo itself.
The Turing test, developed by Alan Turing in 1950, is a test of a machine’s ability to exhibit intelligent behavior equivalent to, or indistinguishable from, that of a human.
In the history of artificial intelligence, an AI winter is a period of reduced funding and interest in artificial intelligence research. The term was coined by analogy to the idea of a nuclear winter. The field has experienced several hype cycles, followed by disappointment and criticism, followed by funding cuts, followed by renewed interest years or decades later.
Artificial neural networks (ANNs) or connectionist systems are computing systems inspired by the biological neural networks that constitute animal brains. Such systems learn (progressively improve performance) to do tasks by considering examples, generally without task-specific programming. For example, in image recognition, they might learn to identify images that contain cats by analyzing example images that have been manually labeled as “cat” or “no cat” and using the results to identify cats in other images. They have found most use in applications difficult to express in a traditional computer algorithm using rule-based programming.
Machine Learning - Irina Higgins (Senior Research scientist at Deep Mind)
AI = the science of making machines smart. It should be a partnership and not a danger to humanity.
Deep Mind is the most distinguished AI company in the world. Its mission is to solve intelligence and use it to make the world a better place.
Operational definition of intelligence: A single algorithm that can learn and adapt to achieve human level performance across the broadest set of taks.
A machine learns through neural networks and reinforcement learning. Reward functions are necessary.
Video games are a good way for agents to learn.
Go is a game which is 3000 years old with 40m players. There are more states of the game than atoms in the Universe. It’s game based on creativity and intuition (19 x 19 is the largest board)
When playing Lee Sedol (Roger Federer of Go), it made an ‘undiscovered move’ = Move 37
Royal Society - Brian Cox presents Science Matters - Machine Learning and Artificial intelligence
How to build a conscious machine - Ron Chrisley
Why is machine consciousness plausible?
- Success of AI/machine learning
- Implied by a general physicalism
“All the physical knowledge about a bat won’t tell you ‘what it’s like to be a bat’” (Nagel)
Easy (easier) problems of consciousness: discrimination of stimuli, focussing of attention, recall of memories
Hard problem - why is the performance of these functions accompanied by experience?
In philosophy and certain models of psychology, qualia (are claimed to be individual instances of subjective, conscious experience).
The term qualia in Latin means “of what sort” or “of what kind” in a specific instance like “what is it like to taste a specific orange, this particular orange now”.
Qualia = a hard property of consciousness
Daniel Dennett is a die-hard physicalist who believes that ‘quail’ should go the same way as ‘phlogiston’ and ‘mermaids’. The term refers to ‘nothing’.
Perhaps a strategy for reconciling physicalism and qualia is not elimination.
Qualia realists = Dualists
Physicalists = Eliminitavists
Legal, ethical and social implications of AI - Lilian Edwards
Transhumanism - an international intellectual movement that aims to transform the human condition by developing and making widely available sophisticated technologies to greatly enhance human intellect and physiology.
The Future of AI - Toby Walsh
We must manage people’s misconceptions and expectations.
Landmark moment for Google when Alpha Go beat the world’s best player at Go.
What AI cannot do:
Learn from a single example (humans can) Act like a human Reason like a human Act autonomously Take over the planet
Do not worry what Elon Musk is tweeting
Concerns over AI are completely overblown
Prime numbers were created by humans: 1,2,3,5,7,11,13,17
Refactorable numbers were created by a machine: 1,2,8,9,12,18,24,36,40
A refactorable number is an integer n that is divisible by the count of its divisors.
Jobs of the future
1/ Technically literate - invent the future
The really innovative idea will come out of Universities, not corporations. The future of AI will be written largely in China.
2/ Social skills - emotional intelligence (computers have not caught up yet)
3/ Jobs we appreciate with a human touch e.g. carpentry
