Machine Learning (84)
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?
-
- 7 min
- The New Republic
- 2020
The narrative of Dr. Timnit Gebru’s termination from Google is inextricably bound with Google’s irresponsible practices with training data for its machine learning algorithms. Using large data sets to train Natural Language Processing algorithms is ultimately a harmful practice because for all the harms to the environment and biases against certain languages it causes, machines still cannot fully comprehend human language.
- The New Republic
- 2020
-
- 7 min
- The New Republic
- 2020
Who Gets a Say in Our Dystopian Tech Future?
The narrative of Dr. Timnit Gebru’s termination from Google is inextricably bound with Google’s irresponsible practices with training data for its machine learning algorithms. Using large data sets to train Natural Language Processing algorithms is ultimately a harmful practice because for all the harms to the environment and biases against certain languages it causes, machines still cannot fully comprehend human language.
Should machines be trusted to handle and process the incredibly nuanced meaning of human language? How do different understandings of what languages and words mean and represent become harmful when a minority of people are deciding how to train NLP algorithms? How do tech monopolies prevent more diverse voices from entering this conversation?
-
- 7 min
- Chronicle
- 2021
The history of AI contains a pendulum which swings back and forth between two approaches to artificial intelligence; symbolic AI, which tries to replicate human reasoning, and neural networks/deep learning, which try to replicate the human brain.
- Chronicle
- 2021
-
- 7 min
- Chronicle
- 2021
Artificial Intelligence Is a House Divided
The history of AI contains a pendulum which swings back and forth between two approaches to artificial intelligence; symbolic AI, which tries to replicate human reasoning, and neural networks/deep learning, which try to replicate the human brain.
Which approach to AI (symbolic or neural networks) do you believe leads to greater transparency? Which approach to AI do you believe might be more effective in accomplishing a certain goal? Does one approach make you feel more comfortable than the other? How could these two approaches be synthesized, if at all?
-
- 7 min
- MIT Tech Review
- 2020
This article examines several case studies from the year of 2020 to discuss the widespread usage, and potential for limitation, of facial recognition technology. The author argues that its potential for training and identification using social media platforms in conjunction with its use by law enforcement is dangerous for minority groups and protestors alike.
- MIT Tech Review
- 2020
-
- 7 min
- MIT Tech Review
- 2020
Why 2020 was a pivotal, contradictory year for facial recognition
This article examines several case studies from the year of 2020 to discuss the widespread usage, and potential for limitation, of facial recognition technology. The author argues that its potential for training and identification using social media platforms in conjunction with its use by law enforcement is dangerous for minority groups and protestors alike.
Should there be a national moratorium on facial recognition technology? How can it be ensured that smaller companies like Clearview AI are more carefully watched and regulated? Do we consent to having or faces identified any time we post something to social media?
-
- 5 min
- Venture Beat
- 2021
Relates the story of Google’s inspection of Margaret Mitchell’s account in the wake of Timnit Gebru’s firing from Google’s AI ethics division. With authorities in AI ethics clearly under fire, the Alphabet Worker’s Union aims to ensure that workers who can ensure ethical perspectives of AI development and deployment.
- Venture Beat
- 2021
-
- 5 min
- Venture Beat
- 2021
Google targets AI ethics lead Margaret Mitchell after firing Timnit Gebru
Relates the story of Google’s inspection of Margaret Mitchell’s account in the wake of Timnit Gebru’s firing from Google’s AI ethics division. With authorities in AI ethics clearly under fire, the Alphabet Worker’s Union aims to ensure that workers who can ensure ethical perspectives of AI development and deployment.
How can bias in tech monopolies be mitigated? How can authorities on AI ethics be positioned in such a way that they cannot be fired when developers do not want to listen to them?
-
- 7 min
- VentureBeat
- 2021
New research and code was released in early 2021 to demonstrate that the training data for Natural Language Processing algorithms is not as robust as it could be. The project, Robustness Gym, allows researchers and computer scientists to approach training data with more scrutiny, organizing this data and testing the results of preliminary runs through the algorithm to see what can be improved upon and how.
- VentureBeat
- 2021
-
- 7 min
- VentureBeat
- 2021
Salesforce researchers release framework to test NLP model robustness
New research and code was released in early 2021 to demonstrate that the training data for Natural Language Processing algorithms is not as robust as it could be. The project, Robustness Gym, allows researchers and computer scientists to approach training data with more scrutiny, organizing this data and testing the results of preliminary runs through the algorithm to see what can be improved upon and how.
What does “robustness” in a natural language processing algorithm mean to you? Should machines always be taught to automatically associate certain words or terms? What are the consequences of large corporations not using the most robust training data for their NLP algorithms?