Transcript PHL 100

Should Artificial Intelligence
be given
Civil Rights
By: Robert Tatoian
Why did i want to research this?
Now a little bit about why I chose this.
There are a few reasons that I wanted to research this topic.
1. There’s not a lot of discussion about giving Artificial
Intelligence actual rights.
2. I’m interested in the field of AI research and know about some
of the topics.
3. In relation to number 1 it’s better to have planned ahead
rather than after the fact.
History of Artificial Intelligence
Research into Artificial Intelligence (AI) was pioneered by John
McCarthy, Marvin Minsky, Allen Newell and Herbert Simon.
The field was officially founded in the 1950’s and until the 1970’s
was funded by the various governments around the world.
It wasn’t until around the 1990’s that AI research was truly refunded and in the same decade many successes with AI
occurred.
Current Developments
When I started writing this a few questions came to mind that
others might have, in addition to my research I’ll also try to
answer these.
1. Just how close are we to creating an AI?
2. Why does this matter to you?
3. They’re just computers why should they have rights?
Current Developments
We are getting closer to creating an Artificial Intelligence that can
pass the “Turning Test.”
The Turning Test was created by Alan Turning
The test works by having a human engage in a text conversation
with both a hidden human and computer, the test is considered
to have succeeded when the human can’t tell the difference
between the two 30% of the time.
Current Developments
The program called “Eugene” was developed by Russian
scientists. Around 2012 this program tricked the humans 29.2%
of the time. Which is .8% away from passing the test.
No other program that I could find has come this close. Not even
IBM’s highly publicized “Watson” computer.
So why is the Turning Test so important?
The basic idea is that if you can’t tell the difference between a
computer or human than the computer is intelligent.
Current Developments
While we are far from creating a truly human intelligence,
programs like “Watson” and “Eugene” represent huge strides
towards that goal.
This is why I think that we should start thinking about whether or
not we should give machines rights. Which ever decided will
have a great impact on the future.
However since AI is still a relatively new field and there have
been no machines that have even passed the Turning test, there
is no real academic discussion about this (at least that I could
find published).
Why you should Care
Two Sides of a Coin
Why should we care about what rights an AI has?
It all depends on what kind of AI it is. You could, in theory,
develop an AI for every task out there, provided it has the means
to execute the task.
Most if not all AI now are specialized for a particular task. We
have to decide if a construct is truly intelligent or if it’s just
following a highly specialized programing.
Two Sides of a Coin
If an AI construct is truly intelligent, then what rights should it
have?
An intelligent construct is hard to define because there is nothing
else like it in the world, aside from people. The definition could
be similar to the definition of a person.
One problem that arises is that all construct have to be
programmed by a person, starting out as nothing but lines of
code. Since it would be nothing more than code some say that it
could never be free to use whatever rights we give it, given that it
could be reprogramed to “think” a different way.
However an argument to that is that humans are
also“programmed” by instinct.
Two sides of a coin
If safeguards were put in
place that prevented
tampering and the
“program” had the ability to
learn, eventually there
would be little to separate
it from humans.
The new AI would be
similar to a baby needing a
“parent” to watch over it.
One example of a AI that is
continually learning and
has a desire to be human
is Data from Star Trek
pictured on right.
Two Sides of a Coin
However if a “construct” is nothing more than a carefully
executed program then by that very admission it doesn’t need
rights and couldn’t do anything with them if it was given them.
Giving a program rights would essentially be the same a giving
your toaster the right to vote and your microwave freedom of
speech.
If any programming is imposed on an AI construct then doesn’t
infringe on free will and wouldn’t that be a violation of any rights
given? However we don’t want our creation to turn on us, so it
could be seen as necessary.
Two sides of a coin
So back to the question why should we care.
We should care because if a “program” can evolve and learn like
a person doesn’t that make it a person?
Keeping something that can reason and think in that capacity in
any type of captivity is immoral and should be illegal.
Finally what reason do we have to not give it rights, if it can
reason and think similarly if not better than a human?
What people have to Say
about aI Rights
views on AI Rights
Some posts on a TED conversation board:
Laszlo Kereszturi: “As regarding the question, the answer is
yes, but only if those computers are like Andrew Martin from
Asimov's "The Bicentennial Man" or "The Positronic Man".
And not like Stephen Byerley from Asimov's "Evidence" or
"The Evitable Conflict".”
Farrukh Yakubov: “Computers are just machines for now. They can never chose
beyond their given capacity. Moreover they are developed and programmed by people,
by engineers and programmers. They lack freewill, even true random number
generator does not exist. All actions of a computer is just consequences of former
states and inputs. Unless you give a computer a human brain, it will always be a
machine, which is the same kind as your desk lamp but a bit complicated. Giving a
right to vote to computer is nonsense. It's like giving a programmer right to vote multiple
times. Computers cannot get human rights only because there is no meaningful
ineffectual entity behind them.”
views on AI Rights
Judge Pau: “First of all, AI is computer but a computer is NOT an AI. ... I had a similar
debate not too long ago in a forum of cyberpunk movie lovers. Don't ask. ... One argue[sic]
that it is the next step of evolution, a post humane era, singularity. I disagree. Is technology
invented to cope with human imperfection or an adventure to celebrate (worship?) human
excellence ? Call me archaic but to me technology however advance[sic] is never meant to
replace human[sic]. They are to complement. An imperfect being creating another life form
to me is mere an act of both irresponsibility and arrogance.”
Finally from a member of our class: “For my personal opinion i really don't think that
Artificial Intelligence would be considered similar to human meaning we wouldn't treat them
the same but they would have some privileges but if they cross some lines i feel like we
would put them to sleep or something...."
How do I answer my own
question
My Answer
Over the course of this project I’ve come to some conclusions.
1. Artificial Intelligence is far away from anything that can think
like a human.
2. There were more factors than I originally thought that would
contribute to whether or not an AI should have rights.
3. There are groups of people who are discussing this albeit
small and there is close to no serious academic discussion
about it.
My Answer
As for whether AI deserve rights I think that there are too many factors that
aren’t determined yet to make an accurate decision.
Both arguments have merit and perhaps there is some balance between the
two.
However based on the current information and articles that I’ve read, I have to
say I think if a computer can process information similarly to a human then it
should be treated similarly to a human.
I think that at moment giving Artificial Intelligence rights seems to be the best
thing to do. I chose this answer not only based on what I’ve read in articles, but
also based on whats going on in the world today. There are plenty of human
rights violations and given the chance people can do some messed up things
to one and other. I think that is would be better if from the start we had laws
about how we would treat our own creations. If we make them capable of
thinking and feeling, then they should also have the same rights that we do.
•
Russell, Stuart J., and Peter Norvig. Artificial Intelligence: A Modern Approach. Upper Saddle River, NJ:
Prentice Hall/Pearson Education, 2003. Print.
Works Cited
•
A Terminator from the Terminator movies. Digital image. ENG 103: Artificial Intelligence. N.p., 26 Aug.
2010. Web. 22 July 2013.
•
Data from Star Trek. Digital image. Wikimedia. N.p., n.d. Web. 22 July 2013.
•
Dilbert Comic. Digital image. N.p., n.d. Web. 22 July 2013.
•
"Progress Report." Interview by Michael Danso. Sakai. University of Rhode Island, 25 June 2013. Web. 22
July 2013. <https://sakai.uri.edu/portal/site/7af563aa-913c-4218-b77a-2de754609a91/page/182207f2bf29-4cc3-9a4e-59605a9f0ff1>.
•
"Russian Brains behind Closest Ever AI Attempt - RT News." Russian Brains behind Closest Ever AI
Attempt - RT News. N.p., 23 Aug. 2012. Web. 22 July 2013.