Introduction to AI

Science

Science (Cognitive Science, Neuroscience, Psychology, Ethology)
The hardest problem in the history of science - How does the brain work?

Engineering

Engineering (Computer Science, AI, Robotics)
The hardest problem in the history of engineering - How can we build an intelligent machine?
Issues:
1. Humans v. children v. animals v. origin of life
2. Design v. learning and evolution
3. Symbolic AI v. Nonsymbolic AI, symbol-grounding
4. Robots v. software-only
5. 1 mind v. multiple minds
6. 1 programmer v. thousands of programmers

Unsolved problems in science

The problem (or the related set of problems) of how the brain, intelligence, consciousness and "free will" work is one of the great unsolved problems of science.

Lists of unsolved problems

Philosophy

Philosophy of Mind (and future science, science fiction)
All the mysterious questions we have thought about for 2500 years that maybe, just maybe, we are going to finally answer.
Issues:
1. Can machines think?
2. Can machines be alive?
3. Are machines already alive?
4. Are we machines?
5. Do we have souls?
6. Does free will exist? (determinism)
7. Do you exist? (homunculi, homunculus argument)

Future Science

Issues:
1. Can we be immortal? (on earth)
2. Can we copy ourselves into new (robot) brains and bodies? (mind uploading, backups, immortality, transhumanism)
3. Can we freeze ourselves now, and wait for that? (cryonics)
4. Can we go to the stars? Or will it be a 2nd species?
5. Will the 2nd species replace us? Is that bad?

Human-machine hybrids

Human-machine hybrids already here (but does not necessarily mean that the sci-fi dreams above are coming):

Human-machine hybrids - cyborgs
Thought controlled machines
- People who are paralysed: thought-controlled wheelchair (simple classification problem)
- People who have lost a limb: neuroprosthetics
- Thought-controlled arm
- Remotely-guided rat
Wikipedia
Buy it yourself:
- OCZ Neural Impulse Actuator

Thought-controlled wheelchair.

Monkey controls a robotic arm with its brainwaves.

The long-term future

Timescale - If AI is solved, if we understand the brain, if we build intelligent machines, when will this happen?

10-50 years? (Imminent)
300 years? (Think of all the changes since 1700. Multiply it.)
1000 years?

Voting Machines Elect One Of Their Own As President

Philosophy and Future of AI - What I think could happen.