Class notes
RLAI - All Themes Lecture Notes
- Course
- Institution
In this file, I have added all my lecture summaries from all Themes of the course. For the exam, I got a 8,6!
[Show more]
Seller
Follow
tigovangerven
Reviews received
Content preview
Robot Law and Artificial IntelligenceLecture 1 - Introduction, Robots, Law
Defining robot
A robot is a man-made object capable of responding to external stimuli without direct human
control.
“A man-made object capable of responding to external stimuli and acting on the world without
requiring direct – some might say constant – human control.”
(required) Elements:
- some sort of sensor or input mechanism (otherwise there are no stimuli to react to);
(camera, mouse)
- some controlling algorithm or other system that will govern the responses to the sensed
data; and
- some ability to respond in a way that affects or at least is noticeable by the world outside
the robot itself.
Examples
1. Sophia
2. Kaspar:
- a small minimally expressive humanoid robot
- it encourages children with low functioning autism to interact with the robot, to break
their isolation and importantly, to facilitate interaction withother people
Crucial elements of robots
1. Physical form
- Excluding software bots
2. Machine / artefact
- Excluding enhanced humans
3. Independency
- No human secretly pulling the strings
Excluding most drones
4. Agency
- What do we mean by agency?
o The ability to do things – take initiative
- How do you know (address agency) – or are you even able to know?
Humans
1. Free will? Agency? (we like to think of ourselves having this)
- Easily manipulated (Kahneman) → not as autonomous as we think; undermines the
concept of free will
2. Emotions?
- Harari: emotions as biological algorithms
- Do robots feel the emotions they express in words → mimic?
, o Kaspar can express human emotions
▪ More a mimic than a real emotion
o How do you know your emotions are true/real?
3. Human dignity?
- How would you describe it, why would we have it?
- Two opposing views: as autonomy or as constraint.
- Nature, automatically - we get this automatically just because we are born human
- A social construction (think about moral status and what being a human entails)
4. Singularity?
a. AI will become smarter than us → what will happen then?
b. Having agency, initiative…
Laws of robotics are referred to in law → but they were fictional and not intended for this
Law
1. Solution for:
- “state of nature”
- Condition humaine
- Shelter, social interaction, sleep, love, support, affection… → set of rules to ensure basic
human requirements
2. Not a coherent, comprehensive system (cant pack everything in “fairly”)
3. Messy, fuzzy, full of contradictions and gaps
4. Aiming for: justice, fairness
a. Subjective notions
5. Implementing moral intuitions
a. Also inconsistent
6. Human rights
7. Produced and enforced mostly by sovereign nation states.
a. Sovereignty – at a specific territory law applies from that territory
b. Scale over effect – if a video is forced off the internet because of one jurisdictions
laws, it will be deleted in all of them, even though its allowed in other jurisdictions
i. Things that happen on the internet are not confined to the territory (state)
ii. Robots – laws are difficult to enforce
So: task for the law
1. Protect human rights and human interests (health!)
a. creates stability in a society
2. Stimulate economy and technological development
a. Always a big argument when talking about regulation
3. Enable cooperation between humans by
a. Aligning expectations
b. Creating order
c. Providing trust
i. No one invests in an economy with no trust, suffering from bribery & fraud
ii. We must cooperate to meet our demands (we need to agree and rely on
others)
iii. Not everything is available in unlimited quantities → need to distribute
resources effectively (some stages are of higher vulnerability → need safety)
iv. Need rules on how we behave towards each other, because of the scarcity of
our resources & allocating them (making ourselves vulnerable sometimes)
v. This cooperation expresses itself in-law/rules all citizens must abide by
, 1. § Primary rules (what you can and cannot do)
a. Coordinate people’s behaviour towards each other
2. § Secondary law (how to enforce and make the law)
a. Coordinate expectations → need to be able to rely on things
3. So that taking the law into your own hands is not necessary (and forbidden to avoid chaos).
Examples of challenges for the law
1. Liability when something goes wrong (self-driving cars)
a. Who/what will be responsible for the damages? – compensation
b. Assigning liability is a form of ensuring safety – provides an incentive for the person
who has insurance on it
2. Algorithmic accountability
a. How to explain “computer says no”
b. How to deal with “Weapons of Math Destruction” (Cathy O’Neill) – computer
systems that decide on us
i. Eg. ecommerce – wealthy neighborhood, you could order and pay after;
versus non-wealthy neighborhood, you had to pay before
ii. More serious: racist, prejudice, discrimination
3. Respect for human dignity and integrity
a. Care robots
b. Sex robots
Examples of challenges for the law (robots)
1. How to treat works created by robots
a. Patent them? Can they receive Nobel prizes?
2. Massive unemployment
a. Universal basic income?
i. Working is a big part of our lives and who we are then?
b. What are we going to do?
3. A right to be un-predicted?
a. Data protection law – eg. look up something and then get ads for the next 2 weeks
about it
i. Freedom and autonomy?
4. Filter bubble / comfort zone (e.g. dystopian scene from Wall-E)
Legal instruments, among others
1. Banning completely
a. E.g. atomic bomb, human cloning, embryo selection
b. International cooperation is crucial
2. Requiring a license
a. So that certain requirements can be set (plus overview plus enforcement)
A. monitored or only used by trained people
3. Stimulating developments
a. E.g. regulatory sandbox, experiments, support research
b. By supporting research you can shift the attention level and obtain more information
on specific topics
4. Liability – insurance
a. E.g. legal personhood of a robot
b. E.g. first party insurance in case of self-driving cars
, c. Someone can be held accountable if a certain invention fails
5. Protection of weaker parties (consumers, users, citizens)
a. E.g. algorithmic decision making
b. This way big firms cannot dominate the world
6. International agreements
a. E.g. killer robots
b. By setting international agreements, one can prohibit certain technologies over the
entire globe in order to achieve worldwide needs and goals
Enforcement is often the biggest challenge!
Lecture 2 - Ethics
Why ethical aspects
1. Regulation on emerging technologies like AI and robots so far only peace-meal
a. No overarching yet
2. Some applications clearly evoke ethical concerns
a. Helps us think about these ethical concerns in a more structured way
b. eg. autonomous cars
3. We all have strong ethical intuitions
a. First sentences of a child: That isn’t fair – we are born with it
b. Animals also have reactions like this (monkey experiment)
c. Envy and fairness
d. Something we are also educated into – natured and nurtured (examples you see
around you)
4. Purpose: help us think about these ethical concerns in a more structured way
a. That feels wrong → not a strong argument
Unclear and inconsistent intuitions → illustrated by six cases
1. Telling a lie
a. Not okay in itself
2. Telling a lie for a very good reason
a. Yes (eg. SS office trying to look for a Jewish person, it's good to lie)
i. Please someone, not offend
ii. Good manners, social structure
3. Harming yourself
a. No
i. drug use, scratching yourself
ii. what about piercings though?
4. Doing a disgusting thing without harming anyone, in the privacy of your own home
a. Yes (eg. buy a deep-fried chicken in a supermarket and have sex with it)
5. Donating to a charity
a. Yes
6. Donating to a charity, for the wrong reason.
a. Yes/Maybe (eg. you want to impress a girl whos asking for a donation)
Two general questions
1. What do we mean when we say “wrong” and “right” in concrete cases?
2. Why are acts wrong (or not) in the cases and in general?
The benefits of buying summaries with Stuvia:
Guaranteed quality through customer reviews
Stuvia customers have reviewed more than 700,000 summaries. This how you know that you are buying the best documents.
Quick and easy check-out
You can quickly pay through credit card or Stuvia-credit for the summaries. There is no membership needed.
Focus on what matters
Your fellow students write the study notes themselves, which is why the documents are always reliable and up-to-date. This ensures you quickly get to the core!
Frequently asked questions
What do I get when I buy this document?
You get a PDF, available immediately after your purchase. The purchased document is accessible anytime, anywhere and indefinitely through your profile.
Satisfaction guarantee: how does it work?
Our satisfaction guarantee ensures that you always find a study document that suits you well. You fill out a form, and our customer service team takes care of the rest.
Who am I buying these notes from?
Stuvia is a marketplace, so you are not buying this document from us, but from seller tigovangerven. Stuvia facilitates payment to the seller.
Will I be stuck with a subscription?
No, you only buy these notes for $9.77. You're not tied to anything after your purchase.
Can Stuvia be trusted?
4.6 stars on Google & Trustpilot (+1000 reviews)
81989 documents were sold in the last 30 days
Founded in 2010, the go-to place to buy study notes for 14 years now