Machine Learning (83)
Find narratives by ethical themes or by technologies.
FILTERreset filters-
- 5 min
- MIT Technology Review
- 2019
In the case of the New Orleans Police Department, along with other cities, data used to train predictive crime algorithms was inconsistent and “dirty” to begin with, making the results disproportionately targeted toward disadvantaged communities.
- MIT Technology Review
- 2019
-
- 5 min
- MIT Technology Review
- 2019
Police across the US are training crime-predicting AIs on falsified data
In the case of the New Orleans Police Department, along with other cities, data used to train predictive crime algorithms was inconsistent and “dirty” to begin with, making the results disproportionately targeted toward disadvantaged communities.
If the data which we train algorithms with is inherently biased, then can we truly ever get a “fair” algorithm? Can AI programs ever solve or remove human bias? What might happen if machines make important criminal justice decisions, such as sentence lengths?
-
- 10 min
- n/a
- 2018
Techniques of misinformation are used to make a film about an alternative history in which the Apollo 11 mission failed and the astronauts became stranded on the moon.
- n/a
- 2018
-
- 10 min
- n/a
- 2018
In Event of Moon Disaster
Techniques of misinformation are used to make a film about an alternative history in which the Apollo 11 mission failed and the astronauts became stranded on the moon.
How do you foresee this type of misinformation being related to conspiracy theories? Do you believe you could have spotted the deepfake were you not specifically looking out for it? Are we approaching a future where we may have to watch all media with such scrutiny?
-
- 15 min
- n/a
- 2018
Choose-your-own-adventure game, in which you experience some sort of data fraud through acting in the position of a cast of characters.
- n/a
- 2018
-
- 15 min
- n/a
- 2018
Choose Your Own Fake News
Choose-your-own-adventure game, in which you experience some sort of data fraud through acting in the position of a cast of characters.
How can you be less vulnerable to fake news and fake advertising online?
-
- 6 min
- n/a
- 2018
Through a series of interactions on a chat and a truth-or-dare type game, the user guesses if they are chatting with a bot or human.
- n/a
- 2018
-
- 6 min
- n/a
- 2018
Bot or Not?
Through a series of interactions on a chat and a truth-or-dare type game, the user guesses if they are chatting with a bot or human.
Are you able to tell the difference between interacting with a bot or human? How? What indicators did you rely on to make your decision?
-
- 30 min
- CNET, New York Times, Gizmodo
- 2023
On May 16, 2023, OpenAI CEO Sam Altman testified in front of Congress on the potential harms of AI and how it ought to be regulated in the future, especially concerning new tools such as ChatGPT and voice imitators.
After watching the CNET video of the top moments from the hearing, read the Gizmodo overview of the hearing and read the associated New York Times article last. All resources highlight the need for governmental intervention to hold companies who generate AI products accountable, especially in the wake of a lack of totally effective congressional action on social media companies. While misinformation and deepfake has been a concern among politicians since the advent of social media, additional new concerns such as a new wave of job loss and crediting artists are raised in the hearing.
- CNET, New York Times, Gizmodo
- 2023
-
- 30 min
- CNET, New York Times, Gizmodo
- 2023
The ChatGPT Congressional Hearing
On May 16, 2023, OpenAI CEO Sam Altman testified in front of Congress on the potential harms of AI and how it ought to be regulated in the future, especially concerning new tools such as ChatGPT and voice imitators.
After watching the CNET video of the top moments from the hearing, read the Gizmodo overview of the hearing and read the associated New York Times article last. All resources highlight the need for governmental intervention to hold companies who generate AI products accountable, especially in the wake of a lack of totally effective congressional action on social media companies. While misinformation and deepfake has been a concern among politicians since the advent of social media, additional new concerns such as a new wave of job loss and crediting artists are raised in the hearing.
If you were in the position of the congresspeople in the hearing, what questions would you ask Sam Altman? Does Sam Altman put too much of the onus of ethical regulation on the government? How would the “license” approach apply to AI companies that already exist/have released popular products? Do you believe Congress might still be able to “meet the moment” on AI?
: Top Moments From ChatGPT Creator's Congressional Testimony