Machine Learning (70)
Find narratives by ethical themes or by technologies.
FILTERreset filters-
- 5 min
- MIT Technology Review
- 2019
Humans take the blame for failures of AI automated systems, protecting the integrity of the technological system and becoming a “liability sponge.” It is necessary to redefine the role of humans in sociotechnical systems.
- MIT Technology Review
- 2019
-
- 5 min
- MIT Technology Review
- 2019
When algorithms mess up, the nearest human gets the blame
Humans take the blame for failures of AI automated systems, protecting the integrity of the technological system and becoming a “liability sponge.” It is necessary to redefine the role of humans in sociotechnical systems.
Should humans take the blame for algorithm-created harm? At what level (development, corporate, or personal) should this liability occur?
-
- 4 min
- VentureBeat
- 2020
A study on the engine of TaskRabbit, an app which uses an algorithm to recommend the best workers for a specific task, demonstrates that even algorithms which attempt to account for fairness and parity in representation can fail to provide what they promise depending on different contexts.
- VentureBeat
- 2020
-
- 4 min
- VentureBeat
- 2020
Researchers Find that Even Fair Hiring Algorithms Can Be Biased
A study on the engine of TaskRabbit, an app which uses an algorithm to recommend the best workers for a specific task, demonstrates that even algorithms which attempt to account for fairness and parity in representation can fail to provide what they promise depending on different contexts.
Can machine learning ever be enacted in a way that fully gets rid of human bias? Is bias encoded into every trained machine learning program? What does the ideal circumstance look like when using digital technologies and machine learning to reach a point of equitable representation in hiring?
-
- 9 min
- Kinolab
- 2002
In the year 2054, the PreCrime police program is about to go national. At PreCrime, three clairvoyant humans known as “PreCogs” are able to forecast future murders by streaming audiovisual data which provides the surrounding details of the crime, including the names of the victims and perpetrators. Although there are no cameras, the implication is that anyone can be under constant surveillance by this program. Once the “algorithm” has gleaned enough data about the future crime, officers move out to stop the murder before it happens. In this narrative, the PreCrime program is audited, and the officers must explain the ethics and philosophies at play behind their systems. After captain John Anderton is accused of a future crime, he flees, and learns of “minority reports,” or instances of disagreement between the Precogs covered up by the department to make the justice system seem infallible.
- Kinolab
- 2002
-
- 9 min
- Kinolab
- 2002
Trusting Machines and Variable Outcomes
In the year 2054, the PreCrime police program is about to go national. At PreCrime, three clairvoyant humans known as “PreCogs” are able to forecast future murders by streaming audiovisual data which provides the surrounding details of the crime, including the names of the victims and perpetrators. Although there are no cameras, the implication is that anyone can be under constant surveillance by this program. Once the “algorithm” has gleaned enough data about the future crime, officers move out to stop the murder before it happens. In this narrative, the PreCrime program is audited, and the officers must explain the ethics and philosophies at play behind their systems. After captain John Anderton is accused of a future crime, he flees, and learns of “minority reports,” or instances of disagreement between the Precogs covered up by the department to make the justice system seem infallible.
What are the problems with taking the results of computer algorithms as infallible or entirely objective? How are such systems prone to bias, especially when two different algorithms might make two different predictions? Is there any way that algorithms could possibly make the justice system more fair? How might humans inflect the results of a predictive crime algorithm in order to serve themselves? Does technology, especially an algorithm such as a crime predictor, need to be made more transparent to its users and the general public so that people do not trust it with a religious sort of fervor?
Minority Report: Predetermination Happens All The Time-
- 3 min
- Kinolab
- 2019
In an imagined future of London, citizens all across the globe are connected to the Feed, a device and network accessed constantly through a brain-computer interface. In this narrative, Tom and Ben find out that their father Lawrence, the creator of the Feed, harvested the Feeds from dead people and used the data stored therein to upload their consciousnesses, including memories and emotions, into a cloud. After seeing the “training data” of Lawrence creating digital consciousnesses on this program, an AI was able to make many more digital consciousnesses of non-real people. These consciousnesses are then able to “possess” human bodies through being uploaded to the Feed devices implanted in real people’s brains.
- Kinolab
- 2019
-
- 3 min
- Kinolab
- 2019
Digitally Reproducing Humans and “Possession”
In an imagined future of London, citizens all across the globe are connected to the Feed, a device and network accessed constantly through a brain-computer interface. In this narrative, Tom and Ben find out that their father Lawrence, the creator of the Feed, harvested the Feeds from dead people and used the data stored therein to upload their consciousnesses, including memories and emotions, into a cloud. After seeing the “training data” of Lawrence creating digital consciousnesses on this program, an AI was able to make many more digital consciousnesses of non-real people. These consciousnesses are then able to “possess” human bodies through being uploaded to the Feed devices implanted in real people’s brains.
Do we as humans need the physical world and our bodies, or can we successfully transfer/upload consciousness and live only in the digital space? Is consciousness the same thing as a soul, or different? Are these people discussed in the clip human, AI, or something in between? What are the far-reaching consequences of AI potentially being able to create realistic consciousnesses? How can brain-computer interfaces implanted into a person lead to a complete loss of their own autonomy? Can and should humans choose to donate their consciousnesses, memories, or emotions to science, with ultimately small knowledge of how these may be deployed?
-
- 13 min
- Kinolab
- 2020
George Almore is an engineer working with a company which hopes to achieve singularity with robots, making their artificial intelligence one step above real humans. In doing this, he works with three prototypes: J1, J2, and J3, each one more advanced than the last. Simultaneously, he plans to upload his dead wife’s consciousness into the J3 robot in order to extend her life. The narrative begins with him explaining his goal to J3 as he has this robot go through taste and emotion tests. Eventually, J3 has evolved into a humanoid robot who takes on the traits of George’s wife, leaving the earlier two versions, who all have a sibling-like bond with each other, feeling neglected.
- Kinolab
- 2020
-
- 13 min
- Kinolab
- 2020
Prototypes, Evolution, and Replacement with Robots
George Almore is an engineer working with a company which hopes to achieve singularity with robots, making their artificial intelligence one step above real humans. In doing this, he works with three prototypes: J1, J2, and J3, each one more advanced than the last. Simultaneously, he plans to upload his dead wife’s consciousness into the J3 robot in order to extend her life. The narrative begins with him explaining his goal to J3 as he has this robot go through taste and emotion tests. Eventually, J3 has evolved into a humanoid robot who takes on the traits of George’s wife, leaving the earlier two versions, who all have a sibling-like bond with each other, feeling neglected.
Are taste and emotion examples of necessary elements of creating advanced AI? If so, why? What good does having these abilities serve in terms of the AI’s relationship to the human world? Is it right to transfer consciousness or elements of consciousness from a deceased person into one or several AI? In the AI, how much is too much similarity to a pre-existing person? Can total similarity ever be achieved, and how? Can advanced AI feel negative human emotions and face mental health problems such as depression? Is it ethical to program AI to feel such emotions, knowing the risks associated with them, including bonding with former or flawed prototypes of itself? If an AI kills itself, does the onus fall on the machine or the human creator?
-
- 12 min
- Kinolab
- 1965
The city of Alphaville is under the complete rule of Alpha-60, an omnipresent robot whose knowledge is more vast than that of any human. This robot, whose learning and knowledge model is deemed “too complex for human understanding,” cements its rule through effectively outlawing emotion in Alphaville, with all definitions of consciousness centering on rationality. All words expressing curiosity or emotion are erased from human access, with asking “why” being replaced with saying “because.” Lemmy is a detective who has entered Alphaville from an external land to destroy Alpha-60. However, in their conversation, Alpha-60 is immediately able to suss out the suspicious aspects of Lemmy’s visit and character.
- Kinolab
- 1965
-
- 12 min
- Kinolab
- 1965
Supercomputer Rule and Condensing Human Behavior
The city of Alphaville is under the complete rule of Alpha-60, an omnipresent robot whose knowledge is more vast than that of any human. This robot, whose learning and knowledge model is deemed “too complex for human understanding,” cements its rule through effectively outlawing emotion in Alphaville, with all definitions of consciousness centering on rationality. All words expressing curiosity or emotion are erased from human access, with asking “why” being replaced with saying “because.” Lemmy is a detective who has entered Alphaville from an external land to destroy Alpha-60. However, in their conversation, Alpha-60 is immediately able to suss out the suspicious aspects of Lemmy’s visit and character.
Can governing computers or machines ever be totally objective? Is this objectivity dangerous? Do humans need emotionality to define themselves and their societies, so that a focus on rationality does not allow for computers to take over rule of our societies? Can the actions of all humans throughout the past be reduced down to a pattern that computers can understand or manipulate? What are the implications of omnipresent technology in city settings?