Promotion of Human Values (161)
Find narratives by ethical themes or by technologies.
FILTERreset filters-
- 7 min
- ZDNet
- 2020
Dr. Gary Marcus explains that deep machine learning as it currently exists is not maximizing the potential of AI to collect and process knowledge. He essentially argues that these machine “brains” should have more innate knowledge than they do, similar to how animal brains function in processing an environment. Ideally, this sort of baseline knowledge would be used to collect and process information from “Knowledge graphs,” a semantic web of information available on the internet which can sometimes be hard for an AI to process without translation to machine vocabularies such as RDF.
- ZDNet
- 2020
-
- 7 min
- ZDNet
- 2020
Rebooting AI: Deep learning, meet knowledge graphs
Dr. Gary Marcus explains that deep machine learning as it currently exists is not maximizing the potential of AI to collect and process knowledge. He essentially argues that these machine “brains” should have more innate knowledge than they do, similar to how animal brains function in processing an environment. Ideally, this sort of baseline knowledge would be used to collect and process information from “Knowledge graphs,” a semantic web of information available on the internet which can sometimes be hard for an AI to process without translation to machine vocabularies such as RDF.
Does giving a machine similar learning capabilities to humans and animals bring artificial intelligence closer to singularity? Should humans ultimately be in control of what a machine learns? What is problematic about leaving AI less capable of understanding semantic webs?
-
- 7 min
- Wired
- 2020
In discussing the history of the singular Internet that many global users experience every day, this article reveals some dangers of digital technologies becoming transparent through repeated use and reliance. Namely, it becomes more difficult to imagine a world where there could be alternatives to the current digital way of doing things.
- Wired
- 2020
-
- 7 min
- Wired
- 2020
Hello, World! It is ‘I’, the Internet
In discussing the history of the singular Internet that many global users experience every day, this article reveals some dangers of digital technologies becoming transparent through repeated use and reliance. Namely, it becomes more difficult to imagine a world where there could be alternatives to the current digital way of doing things.
Is it too late to imagine alternatives to the Internet? How could people be convinced to get on board with a radical redo of the internet as we know it? Do alternatives need to be imagined before forming a certain digital product or service, especially if they end up being as revolutionary as the internet? Are the most popular and powerful digital technologies and services “tools”, or have they reached the status of cultural norms and conduits?
-
- 7 min
- MIT Technology Review
- 2020
This article details a new approach emerging in AI science; instead of using 16 bits to represent pieces of data which train an algorithm, a logarithmic scale can be used to reduce this number to four, which is more efficient in terms of time and energy. This may allow machine learning algorithms to be trained on smartphones, enhancing user privacy. Otherwise, this may not change much in the AI landscape, especially in terms of helping machine learning reach new horizons.
- MIT Technology Review
- 2020
-
- 7 min
- MIT Technology Review
- 2020
Tiny four-bit computers are now all you need to train AI
This article details a new approach emerging in AI science; instead of using 16 bits to represent pieces of data which train an algorithm, a logarithmic scale can be used to reduce this number to four, which is more efficient in terms of time and energy. This may allow machine learning algorithms to be trained on smartphones, enhancing user privacy. Otherwise, this may not change much in the AI landscape, especially in terms of helping machine learning reach new horizons.
Does more efficiency mean more data would be wanted or needed? Would that be a good thing, a bad thing, or potentially both?
-
- 6 min
- TED
- 2020
Jamila Gordon, an AI activist and the CEO and founder of Lumachain, tells her story as a refugee from Ethiopia to illuminate the great strokes of luck that eventually brought her to her important position in the global tech industry. This makes the strong case for introducing AI into the workplace, as approaches using computer vision can lead to greater safety and machine learning can be applied to help those who may speak a language not dominant in that workplace or culture train and acclimate more effectively.
- TED
- 2020
-
- 6 min
- TED
- 2020
How AI can help shatter barriers to equality
Jamila Gordon, an AI activist and the CEO and founder of Lumachain, tells her story as a refugee from Ethiopia to illuminate the great strokes of luck that eventually brought her to her important position in the global tech industry. This makes the strong case for introducing AI into the workplace, as approaches using computer vision can lead to greater safety and machine learning can be applied to help those who may speak a language not dominant in that workplace or culture train and acclimate more effectively.
Would constant computer vision surveillance of a workplace be ultimately positive or negative or both? How could it be ensured that machine learning algorithms were only used for positive forces in a workplace? What responsibility to large companies have to help those in less privileged countries access digital fluency?
How AI can help shatter barriers to equality: Jamila Gordon-
- 5 min
- ZDNet
- 2020
In recent municipal elections in Brazil, the software and hardware of a machine learning technology provided by Oracle failed to properly do its job in counting the votes. This ultimately led to a delay in the results, as the AI had not been properly calibrated beforehand.
- ZDNet
- 2020
-
- 5 min
- ZDNet
- 2020
AI Failure in Elections
In recent municipal elections in Brazil, the software and hardware of a machine learning technology provided by Oracle failed to properly do its job in counting the votes. This ultimately led to a delay in the results, as the AI had not been properly calibrated beforehand.
Who had responsibility to fully test and calibrate this AI before it was used for an election? What sorts of more dire consequences could result from a failure of AI to properly count votes? What are the implications of an American tech monopoly providing this faulty technology to another country’s elections?
-
- 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?