Category: Artificial Intelligence

CIFAR Amii Summer Institute On AI And Society

Last week I attended the CIFAR and Amii Summer Institute on AI and Society. This brought together a group of faculty and new scholars to workshop ideas about AI, Ethics and Society. You can see  conference notes here on philosophi.ca. Some of the interventions that struck me included:

  • Rich Sutton talked about how AI is a way of trying to understand what it is to be human. He defined intelligence as the ability to achieve goals in the world. Reinforcement learning is a form of machine learning configured to achieve autonomous AI and is therefore more ambitious and harder, but also will get us closer to intelligence. RL uses value functions to map states to values; bots then try to maximize rewards (states that have value). It is a simplistic, but powerful idea about intelligence.
  • Jason Millar talked about autonomous vehicles and how right now mobility systems like Google Maps have only one criteria for planning a route for you, namely time to get there. He asked what it would look like to have other criteria like how difficult the driving would be, or the best view, or the least bumps. He wants the mobility systems being developed to be open to different values. These systems will become part of our mobility infrastructure.

After a day of talks, during which I gave a talk about the history of discussions about autonomy, we had a day and a half of workshops where groups formed and developed things. I was part of a team that developed a critique of the EU Guidelines for Trustworthy AI.

Conference notes for CSDH 2019

In early June I was at the Congress for the Humanities and Social Sciences. I took conference notes on the Canadian Society for Digital Humanities 2019 event and on the Canadian Game Studies Association conference, 2019. I was involved in a number of papers:

  • Exploring through Markup: Recovering COCOA. This paper looked at an experimental Voyant tool that allows one to use COCOA markup as a way of exploring a text in different ways. COCOA markup is a simple form of markup that was superseded by XML languages like those developed with the TEI. The paper recovered some of the history of markup and what we may have lost.

  • Designing for Sustainability: Maintaining TAPoR and Methodi.ca. This paper was presented by Holly Pickering and discussed the processes we have set up to maintain TAPoR and Methodi.ca.

  • Our team also had two posters, one on “Generative Ethics: Using AI to Generate” that showed a toy that generates statements about artificial intelligence and ethics. The other, “Discovering Digital Methods: An Exploration of Methodica for Humanists” showed what we are doing with Methodi.ca.

Facebook refused to delete an altered video of Nancy Pelosi. Would the same rule apply to Mark Zuckerberg?

‘Imagine this for a second…’ (2019) from Bill Posters on Vimeo.

A ‘deepfake’ of Zuckerberg was uploaded to Instagram and appears to show him delivering an ominous message

The issue of “deepfakes” is big on the internet after someone posted a slowed down video of Nancy Pelosi to make her look drunk and then, after Facebook didn’t take it down a group posted a fake Zuckerberg video. See  Facebook refused to delete an altered video of Nancy Pelosi. Would the same rule apply to Mark Zuckerberg? This video was created by artists Posters and Howe and is part of a series

While the Pelosi video was a crude hack, the Zuckerberg video used AI technology from Canny AI, a company that has developed tools for replacing dialogue in video (which has legitimate uses in localization of educational content, for example.) The artists provided a voice actor with a script and then the AI trained on existing video of Zuckerberg and that of the voice actor to morph Zuckerberg’s facial movements to match the actor’s.

What is interesting is that the Zuckerberg video is part of an installation called Spectre with a number of deliberate fakes that were exhibited at  a venue associated with the Sheffield Doc|Fest. Spectre, as the name suggests, both suggests how our data can be used to create ghost media of us, but also reminds us playfully of that fictional criminal organization that haunted James Bond. We are now being warned that real, but spectral organizations could haunt our democracy, messing with elections anonymously.

Amazon’s Home Surveillance Company Is Putting Suspected Petty Thieves in its Advertisements

Ring, Amazon’s doorbell company, posted a video of a woman suspected of a crime and asked users to call the cops with information.

VICE has a story about how Amazon’s Home Surveillance Company Is Putting Suspected Petty Thieves in its Advertisements. The story is that Ring took out an ad which showed suspicious behaviour. A woman who is presumably innocent until proven guilty is shown clearly in order to sell more alarm systems. This information came from the police.

Needless to say, it raises ethical issues around community policing. Ring has a “Neighbors” app that lets vigilantes report suspicious behaviour creating a form of digital neighbourhood watch. The article references a Motherboard article that suggests that such digital neighbourhood surveillance can lead to racism.

Beyond creating a “new neighborhood watch,” Amazon and Ring are normalizing the use of video surveillance and pitting neighbors against each other. Chris Gilliard, a professor of English at Macomb Community College who studies institutional tech policy, told Motherboard in a phone call that such a “crime and safety” focused platforms can actively reinforce racism.

All we need now is for there to be AI in the mix. Face recognition so you can identify anyone walking past your door.

AI, Ethics And Society

Last week we held a conference on AI, Ethics and Society at the University of Alberta. As I often do, I kept conference notes at: philosophi.ca : AI Ethics And Society.

The conference was opened by Reuben Quinn whose grandfather signed Treaty 6. He challenged us to think about what labels and labelling mean. Later Kim Tallbear challenged us to think about how we want the encounter with other intelligences to go. We don’t have a good track record of encountering the other and respecting intelligence. Now is the time to think about our positionality and to develop protocols for encounters. We should also be open to different forms of intelligence, not just ours.

$432 000 painting “by AI” sold at Christie’s

A painting created using GANs (generative adversarial networks) sold for $432 000 at Christies today.

Last year a $432 000 painting “by AI” sold at Christie’s. The painting was created by a collective called Obvious. They used a Generative Adversarial Network. In an essay titled, A naive yet educated perspective on Art and Artificial Intelligence, they talk about how they created the work.

Generative Adversarial Networks (GANs) analyze tens of thousands of images, learn from their features, and are trained with the aim to create new images that are undistinguishable from the original data source.

They also point out that many of the same concerns people have about AI art today were voiced about photography in the 19th century. Photography automated the image making business much as AIs are automating other tasks.

Can we use these GANs for other generative scholarship?

Centrelink scandal

Data shows 7,456 debts were reduced to zero and another 12,524 partially reduced between July last year and March

The Guardian has a number of stories on the Australian Centrelink scandal including, Centrelink scandal: tens of thousands of welfare debts wiped or reduced. The scandal arose when the government introduce changes to a system for calculating overpayment to welfare recipients and clawing it back that removed a lot of the human oversight. The result was lots of miscalculated debts being automatically assigned to some of the most vulnerable. A report, Paying the Price of Welfare Reform, concluded that,

The research concludes that although welfare reform may be leading to cost savings for the Department of Human Services (DHS), substantial costs are being shifted to vulnerable customers and the community services that support them. It is they that are paying the price of welfare reform.

Continue reading Centrelink scandal

We Built a (Legal) Facial Recognition Machine for $60

The law has not caught up. In the United States, the use of facial recognition is almost wholly unregulated.

The New York Times has an opinion piece by Sahil Chinoy on how (they) We Built a (Legal) Facial Recognition Machine for $60. They describe an inexpensive experiment they ran where they took footage of people walking past some cameras installed in Bryant Park and compared them to known people who work in the area (scraped from web sites of organizations that have offices in the neighborhood.) Everything they did used public resources that others could use. The cameras stream their footage here. Anyone can scrape the images. The image database they gathered came from public web sites. The software is a service (Amazon’s Rekognition?) The article asks us to imagine the resources available to law enforcement.

I’m intrigued by how this experiment by the New York Times. It is a form of design thinking where they have designed something to help us understand the implications of a technology rather than just writing about what others say. Or we could say it is a form of journalistic experimentation.

Why does facial recognition spook us? Is recognizing people something we feel is deeply human? Or is it the potential for recognition in all sorts of situations. Do we need to start guarding our faces?

Facial recognition is categorically different from other forms of surveillance, Mr. Hartzog said, and uniquely dangerous. Faces are hard to hide and can be observed from far away, unlike a fingerprint. Name and face databases of law-abiding citizens, like driver’s license records, already exist. And for the most part, facial recognition surveillance can be set up using cameras already on the streets.

This is one of a number of excellent articles by the New York Times that is part of their Privacy Project.

Are Robots Competing for Your Job?

Are robots competing for your job?
Probably, but don’t count yourself out.

The New Yorker magazine has a great essay by Jill Lepore about whether Are Robots Competing for Your Job? (Feb. 25, 2019) The essay talks about the various predictions, including the prediction that R.I. (Remote Intelligence or global workers) will take your job too. The fear of robots is the other side of the coin of the fear of immigrants which raises questions about why we are panicking over jobs when unemployment is so low.

Misery likes a scapegoat: heads, blame machines; tails, foreigners. But is the present alarm warranted? Panic is not evidence of danger; it’s evidence of panic. Stoking fear of invading robots and of invading immigrants has been going on for a long time, and the predictions of disaster have, generally, been bananas. Oh, but this time it’s different, the robotomizers insist.

Lepore points out how many job categories have been lost only to be replaced by others which is why economists are apparently dismissive of the anxiety.

Some questions we should be asking include:

  • Who benefits from all these warnings about job loss?
  • How do these warnings function rhetorically? What else might they be saying? How are they interpretations of the past by futurists?
  • How is the panic about job losses tied to worries about immigration?

Artificial intelligence: Commission takes forward its work on ethics guidelines

The European Commission has announced the next step in its Artificial Intelligence strategy. See Artificial intelligence: Commission takes forward its work on ethics guidelines. The appointed a High-Level Expert Group in June of 2018. This group has now developed Seven essentials for achieving trustworthy AI:

Trustworthy AI should respect all applicable laws and regulations, as well as a series of requirements; specific assessment lists aim to help verify the application of each of the key requirements:

  • Human agency and oversight: AI systems should enable equitable societies by supporting human agency and fundamental rights, and not decrease, limit or misguide human autonomy.
  • Robustness and safety: Trustworthy AI requires algorithms to be secure, reliable and robust enough to deal with errors or inconsistencies during all life cycle phases of AI systems.
  • Privacy and data governance: Citizens should have full control over their own data, while data concerning them will not be used to harm or discriminate against them.
  • Transparency: The traceability of AI systems should be ensured.
  • Diversity, non-discrimination and fairness: AI systems should consider the whole range of human abilities, skills and requirements, and ensure accessibility.
  • Societal and environmental well-being: AI systems should be used to enhance positive social change and enhance sustainability and ecological responsibility.
  • Accountability: Mechanisms should be put in place to ensure responsibility and accountability for AI systems and their outcomes.

The next step has now been announced and that is a pilot phase that tests these essentials with stakeholders. The Commission also wants to cooperate with “like-minded partners” like Canada.

What would it mean to participate in the pilot?