Philosophy

1
Philosophy AI

• Chapter 26
• Oliver Sacks, The man who mistook his wife for a
  hat

2
Some of the big questions

• What is AI, really?
• What does an intelligent system look like?
• Does an AI needand can it haveemotions,
  consciousness, empathy, love?
• Can we ever achieve AI, even in principle?
• How will we know if weve done it?
• If we can do it, should we?

3
Weak Strong AI

• Weak AI hypothesis
• Machines can be created that will act as if they
  are intelligent
• What AI can do is to develop useful, powerful
  applications
• Strong AI hypothesis
• Machines can be created that will actually think
  (rather than simulate thinking)

4
Weak AI Turing Test

• Basic test
• Interrogator in one room, human in another,
  system in a third
• Interrogator asks questions human and system
  answer
• Interrogator tries to guess which is which
• If the system wins, its passed the Turing Test
• Features of a human-sounding conversational
  partner?
• Problems from a scientific point of view Turing
  test not reproducible, constructive, or amenable
  to mathematical analysis

5
Turing test

• Turing predicted by 2000, 30 chance for a
  machine to fool a lay person for 5 minutes
• Anticipated all major arguments against AI (see
  below)
• Suggested major components of AI language
  understanding, reasoning, knowledge, learning,

6
Turing Test

• An early success at fooling some people Eliza,
  simulating a Rogerian therapist
• Try it! http//www.manifestation.com/neurotoys/eli
  za.php3
• Sample interaction
• Patient Men are all alike.
• ELIZA In what way?
• Patient They are always bugging about something
  or the other
• ELIZA Can you think of a specific example?
• Patient Well, my boyfriend made me come here.
• ELIZA Your boyfriend made you come here?
• Patient He says Im depressed much of the time.
• ELIZA I am sorry you are depressed.

7

• .Eliza the importance of picking the correct
  application area
• Rogerian therapy the therapist does not
  disagree or point out contradictions (Shaffer,
  1978). Neither do they attempt to delve into the
  unconscious.
• A psychology joke
• Patient to Rogerian therapist I'm really
  depressed.
• Therapist I see. Yes. You are depressed.
• Patient Nothing is going well.
• Therapist Nothing well.
• Patient I feel like killing myself.
• Therapist You're thinking of killing yourself.
• Patient Yes, I'm going to do it NOW.
• Therapist You want to do it now.
• Patient Jumps out window.
• Therapist Woosh. Splat.

8
Objections to the Turing Test

• Argument from disability hey, there are lots
  of things that a computer cant do!
• But now there are lots of things that a computer
  can do better than us which tasks are the
  significant ones?
• Mathematical objection some mathematical
  questions are unanswerable by particular formal
  systems
• Are these questions significant to judging
  intelligence?
• Are computers Turing machines? (Turing machines
  are infinite, computers are finite)
• What happens if we apply these arguments to
  people?

9
objections

• Argument from informality what people do is
  too complex to capture
• Progress in incorporating background knowledge
• Learning algorithms increasing ability to
  operate autonomously (unsupervised learning),
  learn new features, prune feature spaces
• The chess example human experts and software
  may use different techniques--or do they?

10
Strong AI can machines really think?

• Objections
• Theological objections
• Machines cant be conscious or feel emotions
• Reductionism doesnt really answer the question
  why cant machines be conscious or feel
  emotions??
• Turing we cant prove that other people are
  conscious so why do we have a higher standard
  for computers?
• Machines dont have Human Quality X
• Machines just do what we tell them to do
• Maybe people just do what their neurons tell them
  to do

11
objections

• brain in a vat experiment
• Is physicality crucial for intelligence?
• Matrix scenario a brain is supported, bodiless,
  in a vat, and signals simulating a virtual world
  are fed in/out of the brain
• Is being hungry the same as
• DyingFor (Me, Pizza)

12
objections

• The Chinese Room John Searle, 1980
• Human who knows only English stacks of paper
  with Chinese symbols rule book in English,
  stating which bit of paper to give in response to
  a given (Chinese) input
• Human who knows only Chinese on outside of room
  passes in Chinese query, receives Chinese
  response

13

• The Chinese Room
• understanding is necessary to intelligence,
  and the Chinese Room doesnt have it
• What is understanding?
• How do we know people have it?
• If a room doesnt have it, does that mean a
  computer cant?
• Reasoning from intuition--not provable
• Back to brains cause minds
• I used to spend a good deal of time in
  comp.ai.philosophy arguing about Searle, and no
  one ever budged an inch. It was the Esperanto or
  the Middle East crisis of philosophy. Worse,
  really-- I've known people who've reversed their
  politics or abandoned an auxlang, but it seems
  that once someone has formed an opinion on the
  Chinese Room, it's permanent.
• --Mark Rosenfelder

14
objections

• Computers cant be creative
• Two paintings produced by Harold Cohens Aaron
  software

15
Ethics Risks of AI

• People might lose their jobs to automation
• Regular computing lower paid workers lose their
  jobs (bookkeepers, typists, )
• AI potential for computers to do higher paid,
  interesting jobs (think expert systems for
  medical diagnosis much earlier, automated
  programming replacing programmers)
• So far AI does work that people cant do
  because of cost (spam filters fraud detection in
  credit card transactions)

16
ethics

• People might have too much/little leisure
• Maybe automated agents can take up some of the
  work, and give us more free time!
• Or maybe the agents can bring us more
  information/work, and give us less time.
• there is a large reward for being slightly
  better than the competition working 10 more
  could mean a 100 increase in income

17
ethics

• People might lose their sense of being unique
• If an AI is created, wont that also mean that
  people are equivalent to automata? Will we lose
  our humanity?
• but weve managed to lose the idea of being the
  center of the universe (Copernicus) and idea of
  being separate from nature (Darwin) without
  changing much

18
ethics

• People might lose their privacy
• intelligent scanning of electronic text,
  telephone conversations, recorded conversations
• Machine learning data matching
• AI systems would lead to loss of accountability
• If an expert medical diagnosis system exists, and
  kills a patient with an incorrect diagnosis, who
  do we sue?
• The end of the human race! Robots take over the
  world!!

19
AI state of the art are we there yet?

• Which of the following can be done at present?
• Play a decent game of table tennis
• Drive safely along a curving moutain road
• Drive safely along a straight empty highway
• Buy a weeks worth of groceries on the web
• Buy a weeks worth of groceries at PaknSave
• Play a decent game of bridge
• Discover and prove a new mathematical theorem

20
not quite

• Which of the following can be done at present?
• Design and execute a research program in
  molecular biology
• Write an intentionally funny story
• Give competent legal advice in a specialised area
  of law (housing, employment, tax, )
• Translate spoken English into spoken Swedish in
  realtime
• Converse successfully with another person for an
  hour
• Perform a complex surgical operation
• Unload any dishwasher and put everything away