FFAI Objections to AI

Get Started. It's Free
or sign up with your email address
Rocket clouds
FFAI Objections to AI by Mind Map: FFAI Objections to AI

1. common objections to AI

1.1. instrinsic inabilities

1.1.1. feel emotions

1.1.2. feel love

1.1.3. self-awareness

1.1.4. be religious

1.2. limited computational capabilities

1.2.1. hypercomputational minds

1.2.2. all computers are Turing equivalent

1.2.3. humans can solve problems that a Turing machine can't solve, so humans are more powerful than machines

1.3. physicality

1.3.1. you need a human body to be "intelligent"

1.3.2. two criteria

1.3.2.1. you really need a full human body

1.3.2.2. a human brain suffices, it's the biological neural hardware that makes the difference

1.3.3. two possible meanings

1.3.3.1. that is our definition because we don't want to include robots (cf definition of "planet")

1.3.3.2. physicality with a biological body is necessary for anything resembling intelligence

1.4. computers only follow rules

1.4.1. therefore, they can't innovate

1.4.2. therefore, they can't deal with complicated problems requiring human judgment

1.5. computers don't have "consciousness"

1.5.1. they may act like they do, but they actually don't

1.5.2. what does this mean? is it like sleepwalking?

1.5.3. "texting while sleeping"?

1.6. computers don't have "free will"

1.6.1. we'll have a separate lecture on that

1.7. computers don't have "qualia"

1.7.1. humans have an experience when seeing the color "red"

1.7.2. robots lack that internal experience; they only manipulate symbols

1.8. computers don't have a "soul"

1.8.1. dualism

1.8.2. there is a separate thing that we call the "soul"

1.8.3. the soul is connected to the brain somehow

1.8.4. humans have this soul, no other physical system does

1.8.5. but which humans? and when? do you know what you religion says about that?

1.9. computers don't have "self-awareness"

1.9.1. self-awareness means the ability to recognize that you are an agent in the world just like other agents

1.9.2. ... and you can use that knowledge to succeed better in the world

1.10. computers lack "morality"

1.10.1. what about sociopaths?

2. Claim: computers may be able to simulate intelligence, but they can't really be intelligent in a human sense.

3. (This is our first exploration of these topics; we'll have more to say about "consciousness", "free will", etc. later)

4. think about

4.1. How do you know someone else does?

4.2. How do you know you yourself do?

4.3. What experiments could you do to tell?

4.4. If you were a robot, would you even know?

4.5. All of these (and more) have been explored endlessly in movies.

4.5.1. Can you name examples?

4.5.2. Have a look at TV tropes (tvtropes.org)

5. kinds of AI

5.1. strong AI

5.1.1. the computer "is human"

5.1.1.1. sapience - it is smart

5.1.1.2. sentience - experiences pain and pleasure

5.1.2. i.e. cyborg (brain in a robot body) and AI (intelligent computer in a robot body) are "the same"

5.1.3. killing it is murder

5.2. weak AI

5.2.1. the computer "acts human"

5.2.1.1. sapience - it is smart

5.2.1.2. no sentience - it doesn't experience pain and pleasure, although it may try to pretend it does

5.2.2. but it may not actually "be" human

5.2.3. it's a "simulation"