User:Rachelkmoy/sandbox

Polish Your Work

Resurrecting Late Personas

Deepfakes technology can not only be used to fabricate messages and actions of others, but it can also be used to revive individuals who have passed away. On October 29, 2020, Kim Kardashian posted a video of her late father Robert Kardashian; the face in the video of Robert Kardashian was created with deepfake technogy. This hologram was created by the company Kaleida, where they use a combination of performance, motion tracking, SFX, VFX and DeepFake technologies in their hologram creation.

There was also an instance where Joaquin Oliver, victim of the Parkland shooting was resurrected with deepfake technology. Oliver's parents teamed up on behalf of their organization Nonprofit Change the Ref, with McCann Health to produce this deepfake video advocating for gun-safety voting campaign. In this deepfake message, it shows Joaquin encouraging viewers to vote.

Continue improving your article

Images:

Draft Suggestion: add images

• I researched and added an image of Doug Roble's deepfake debuted at his TED talk that complied with Wikipedia's proper licensing.
• I researched and added an image that depicts the facial morphing and swapping that artificial intelligence uses in the creation of deepfakes, also making sure it complied with Wikipedia's image licensing.

Politics and Deepfakes

I researched examples of deepfakes being created of politicians from both political parties in order to make make sure there is no bias.

Donald Trump Deepfake

A deepfake of Donald Trump was easily created based on a skit Jimmy Fallon performed on NBC’s The Tonight Show. In this skit (aired May 4, 2016), Jimmy Fallon dressed up as Donald Trump and pretended to participate in a phone call with Barack Obama, conversing in a manner that presented him to be bragging about his primary win in Indiana. On May 5, 2019 a deepfake of Donald Trump (taken from this skit) was created. In this deepfake, Jimmy Fallon's face was transformed into Donald Trump's face (audio remained the same). This deepfake video was uploaded to YouTube by the founder of Derpfakes with a comedic intent.

Barack Obama Deepfake

American actor Jordan Peele, BuzzFeed, and Monkeypaw Productions created and produced a deepfake of Barack Obama (uploaded to YouTube on Apr 17, 2018) that depicted Barack Obama cursing and calling Donald Trump names. In this deepfake Peele's voice and mouth was transformed and manipulated into Obama's voice and face. The intent for this video was to portray the dangerous consequences and power of deepfakes, and how deepfakes can make anyone say anything.

Lead Section

When I add my contributions to the article, I plan to reorganize the lead section to make sure that my contributions merge with the existing information seamlessly. For example, since there is a great amount of information surrounding the involvement of deepfakes in politics, I will create a section dedicated to this instead of having an imbalance of information where "Politics" is a subcategory.

Positive Implications of Deepfakes

Positive and beneficial innovations have also emerged alongside the growing popularity and creation of deepfakes. For example, corporate training videos can be created using deepfaked avatars and their voices. An example of this is Synthesia, which uses deepfake technology with avatars to create personalized videos. This use of deepfakes allow companies to decrease cost and time spent on training employees.

Microsoft has developed an app called "Seeing AI" which uses artificial intelligence and deepfake technology to narrate the world around you to aid individuals you are blind and/or have low vision. With deepfake technology, this app can narrate text in documents, scan products and barcodes, recognize people and their emotions, describe the location and setting around you, identify currency and bills, and communicate these features in different languages in a volume and tone that is adjusted to your surroundings.

Summary of Revisions

To further improve and revise my article, I have added images with relevant captions, updated the lead section, researched and added examples of events of deepfakes from both political parties to combat potential political bias, and added a section highlighting the positive impacts from deepfakes in order to balance the tone surrounding deepfakes in this article. In addition to this, I have checked my grammar and spacing in my draft to ensure that there are no errors and mistakes. I believe that these sections that I chose to revise are the most significant and important to improve from my response in "Response to peer reviews."

Facial Morphing

Deepfake of Doug Roble, "DigiDoug" debuted at TED2019

Response to peer reviews

Response to Dalexandertom:

Response to Lead Feedback: There is no suggestions on improving the lead section as it is concise and relevant to the content.

Response to Content Feedback: There is no suggestions on improving the content section as the information included is up to date, relevant, and not biased.

Response to Tone and Balance: I will make sure to research and locate relevant evidence/support for both political parties so that there is not any biased tone in favor of a particular side.

Response to Sources and Reference: There is no suggestions on improving the sources and references as they are reliable sources, and the links work.

Response to Organization: There is no suggestions on improving the organization as the content is presented in an easily understood manner with no grammatical errors.

Responses to Images and Media: I will make sure to add images and relevant captions to it.

Response to For New Articles Only: There seems to be no suggestions on improving The News Article impressions as it adds valuable information to the article topic.

Response to Brandonqin

Response to Lead Feedback: I will make sure to improve the lead section to include the topic of government and political usage, and how it touches upon social media usage, fraud, and current news.

Response to Content Feedback: I will make sure to include a diverse range of examples of deepfakes affecting both political parties so there is no bias.

Response to Tone and Balance: I will make sure to research and locate relevant evidence/support for both political parties so that there is not any biased tone in favor of a particular side. I may also look into other platforms that fall into the extremes of both political parties, and how that may affect its audiences.

Response to Sources and Reference: There is no suggestions on improving the sources and references as they are reliable sources, and the links work. I may also take a look into Wikipedia's marginalized individual goals.

Response to Organization: I will double check the spacing of my article and any underlined words/phrases to see if it is necessary/grammatically correct.

Response to Kyle.chan201

(For some reason, when I click on the peer review there is nothing there—mistake possibly?)

Response to Esk00

(For some reason, when I click on the peer review there is nothing there—mistake possibly?)

Response to Ethanpak

(For some reason, when I click on the peer review there is nothing there—mistake possibly?)

Response to RiaVora

Response to Lead Feedback: There is no suggestions on improving the lead section as it is concise and relevant to the content.

Response to Content Feedback: I will look into researching specific quotes to implement (for example, Twitter policies).

Response to Tone and Balance: I will make sure to research and locate relevant evidence/support for deepfakes so the article does not seem biased towards deepfakes.

Response to Sources and Reference: I will make sure to research if there are more sources to back up the same points so that points are not backed up by the same sources multiple times. I will also research and find sources outside of policies from direct social media platforms so there is no bias.

Response to Organization: I will make sure to reword some sentences to make sure that it flows well and isn't choppy. I will also consider creating a separate section for Facebook do to its great history and background to this topic.

Response to For New Articles Only: I will look into more extreme views to incorporate and include multiple different perspectives. I will also look into linking other relevant articles so that it is more discoverable.

Concluding Plans for Revision:

I will make sure to research and locate relevant evidence/support for both political parties so that there is not any biased tone in favor of a particular side. I will make sure to add images and relevant captions to it. I will make sure to improve the lead section to include the topic of government and political usage, and how it touches upon social media usage, fraud, and current news. I will make sure to include a diverse range of examples of deepfakes affecting both political parties so there is no bias. I will make sure to research and locate relevant evidence/support for both political parties so that there is not any biased tone in favor of a particular side. I may also look into other platforms that fall into the extremes of both political parties, and how that may affect its audiences. I will double check the spacing of my article and any underlined words/phrases to see if it is necessary/grammatically correct. I will look into researching specific quotes to implement (for example, Twitter policies). I will make sure to research and locate relevant evidence/support for deepfakes so the article does not seem biased towards deepfakes. I will make sure to research if there are more sources to back up the same points so that points are not backed up by the same sources multiple times. I will also research and find sources outside of policies from direct social media platforms so there is no bias. I will make sure to reword some sentences to make sure that it flows well and isn't choppy. I will also consider creating a separate section for Facebook do to its great history and background to this topic. I will look into more extreme views to incorporate and include multiple different perspectives. I will also look into linking other relevant articles so that it is more discoverable.

Start Drafting Your Contributions

What is Missing from the Article:

• Recent regulations
• Recent Advancements
• Implications of Viral Deepfakes
  • Examples

Draft:

Responses from Platforms (Paragraphs 3/5)

Twitter is taking active measures to handle synthetic and manipulated media on their platform. In order to prevent disinformation from spreading, Twitter is placing a notice on tweets that contain manipulated media and/or deepfakes that signal to viewers that the media is manipulated. There will also be a warning that appears to users who plan on retweeting, liking, or engaging with the tweet. Twitter will also work to provide users a link next to the tweet containing manipulated or synthetic media that links to a Twitter Moment or credible news article on the related topic—as a debunking action. Twitter also has the ability to remove any tweets containing deepfakes or manipulated media that may pose a harm to users’ safety. In order to better improve Twitter's detection of deepfakes and manipulated media, Twitter has asked users who are interesting in partnering with them to work on deepfake detection solutions to fill out a form (that is due November 27, 2020).

Facebook has taken efforts towards encouraging the creation of deepfakes in order to develop state of the art deepfake detection software. Facebook was the prominent partner in hosting the Deepfake Detection Challenge (DFDC), held December 2019, to 2114 participants who generated more than 35000 models. The top performing models with the highest detection accuracy were analyzed for similarities and differences; these findings are areas of interest in further research to improve and refine deepfake detection models . Facebook has also detailed that the platform will be taking down media generated with artificial intelligence used to alter an individual's speech. However, media that has been edited to alter the order or context of words in one's message would remain on the site but be labeled as false, since it was not generated by artificial intelligence.

The Defense Advanced Research Projects Agency (DARPA) has funded a project where individuals will compete to create AI-generated videos, audio, and images as well as automated tools to detect these deepfakes. DARPA has even taken efforts to host a "proposers day" for a project affiliated with the Semantic Forensics Program where researchers are driven to prevent viral spread of AI-manipulated media. DARPA and the Semantic Forensics Program are also working together to detect these AI-manipulated media through efforts focused in training computers to utilize common sense, logical reasoning. DARPA has also created a Media Forensics (MediFor) program, to mitigate the increasing harm that deepfakes and AI-generated media poses. This program aims to not only detect deepfakes, but also provide information regarding how the media was created. Simultaneously, DARPA's goal is to address and emphasize the consequential role of deepfakes and their influence upon decision making.

Uses of Deepfakes (Paragraph 4/5)

As deepfake technology increasingly advances, Disney has perfected their visual effects using high resolution deepfake face swapping technology. Disney perfected their technology through progressive training programmed to identify facial expressions, implementing a face swapping feature, and iterating in order to stabilize and refine the output. This high resolution deepfake technology will be implemented into movie and television production—saving significant operational and production costs. Disney's deepfake generation model can produce AI-generated media at a 1024 x 1024 resolution, which is much greater and produces more realistic results than common models that produce media at a 256 x 256 resolution. In addition, with this technology, Disney has the opportunity to revive dead actors and characters with a quick and simpleface swap; Disney can now resurrect and revive characters for fans to enjoy.

Deepfakes and the Politics (Paragraph 5/5)

The consequences of a deepfake is not significant enough to destabilize the entire government system; however, deepfakes possess the ability to damage individual entities tremendously. This is because deepfakes are often targeted at one individual, and/or their relations to others in hopes to create a narrative powerful enough to influence public opinion or beliefs. This can be done through deepfake voice phishing, which manipulates audio to create fake phone calls or conversations. Another method of deepfake use is fabricated private remarks, which manipulated media to convey individuals voicing damaging comments.

Deepfakes of North Korean leader Kim Jong-un and Russian president Vladimir Putin have also been created by a nonpartisan advocacy group RepresentUs. These deepfakes were meant to air publicly as commercials to relay the notion that interferences by these leaders in US elections will be detrimental to the United State's democracy; the commercial also aimed to shock Americans to realize how fragile the democracy is, and how media and news can significantly influence the country's path regardless of credibility. However, those these commercials did include an ending comment detailing that the footage was not real, the commercials ultimately did not air due to fears and sensitivity regarding how Americans may react. Several deepfakes of politicians have already been created and gone viral.

Nancy Pelosi

A clip from Nancy Pelosi's speech at the Center for American Progress given on May 22, 2019 was slowed down, in addition to the pitch being altered, to make it seem as if she were drunk; however, critics argue that this is not a deepfake.

Peer editor: Daniel Tom - Hi Rachel, your contributions are very well thought out and organized. It is a good read and flows nicely. The sources you utilize are very solid and academic. However, I feel like the balance of your article is a little lacking. If I had to make some changes, I would add some additional sources of democratic deep fakes to improve the balance of the article.

Finalize your topic/Find your sources

Selected Article: Deepfake

What can you add? Ideas:

I plan on contributing ideas and media such as the following to my selected article. I think that it will be beneficial as it mentions ideas that were not included in the original article (such as identity theft, solutions..). I think it will also be good because these articles/news are updated, and it is important to update the article as deepfakes are constantly advancing and new solutions, events, and implications of it only continue to grow.

• Updated regulations concerning deepfakes
  • Companies like Instagram , Twitter, US Department of Defense
• Advancements in the technology
• major events/ occurrences where deepfakes were trending
• the possible damage it can impose, election, celebrity reputation..
• pictures, videos, gifs to illustrate the technology

Relevant, reliable books, journal articles, or other sources:

Articles

• https://blog.twitter.com/en_us/topics/company/2019/synthetic_manipulated_media_policy_feedback.html
  • Twitter's action regarding manipulated media
• https://ai.facebook.com/blog/deepfake-detection-challenge-results-an-open-initiative-to-advance-ai/
  • Facebook's Deepfake Challenge to advance AI and create solutions to detect deepfakes
• https://www.technologyreview.com/2018/05/23/142770/the-us-military-is-funding-an-effort-to-catch-deepfakes-and-other-ai-trickery/
  • DARPA's attention to deepfakes

Journals & Publications

• https://studios.disneyresearch.com/2020/06/29/high-resolution-neural-face-swapping-for-visual-effects/
  • Disney's Research efforts towards high resolution face swapping
• https://www.jstor.org/stable/resrep25783.6?seq=1#metadata_info_tab_contents
  • How deepfakes may affect the financial system, how big is the threat, how easy is it to combat, identity theft

Link to my Contribution to the Article's Talk page:

Talk:Deepfake#Some Ideas to Add to the Article, & Sources

Possible Article Topics

Option 1

Article title

Deepfake

Article Evaluation

Seems comprehensive, but the information could be updated with recent regulations and advancements, such as from Twitter, Facebook, Disney, and more.

Is the article's content relevant to the topic? Yes Is it written neutrally? Yes Does each claim have a citation? Yes Are the citations reliable? Yes Does the article tackle one of Wikipedia's equity gaps? No

Sources

https://blog.twitter.com/en_us/topics/company/2019/synthetic_manipulated_media_policy_feedback.html <-- Twitter's updates

https://www.nytimes.com/2020/01/07/technology/facebook-says-it-will-ban-deepfakes.html <-- Facebook's updates

https://www.reviewgeek.com/46711/disneys-face-swapping-technology-is-the-most-convincing-yet/ <-- Disney's increasingly advanced deepfake technology

Option 2

Article title

Fortnite

Article Evaluation

Relates to Civic Tech in a sense that the growth and impact on society in terms of impact, retention, acquisition, and pull, which can be translated to technology's impacts on civic activities. Article could be updated with more insight behind the background and the history of the game, including it's original debut as a $40 game to CEO of Epic Game's Tim Sweeney's efforts in growing the game's user retention and engagement.

Is the article's content relevant to the topic? Yes Is it written neutrally? Yes Does each claim have a citation? Yes Are the citations reliable? Yes Does the article tackle one of Wikipedia's equity gaps? No

Sources

https://www.gallantway.com.au/post/case-study-heres-why-fortnite-may-be-the-ultimate-growth-marketing-example <-- Explains Fortnite's growth in consumer retention, engagement, and acquisition.

https://www.bloomberg.com/news/features/2018-07-24/fortnite-phenomenon-turns-epic-game-developer-into-billionaire <-- Explains significant events behind Fortnite's growth

https://www.wsj.com/articles/wheres-the-off-switch-videogame-developers-are-making-it-harder-to-stop-playing-1534757400 <-- Provides context regarding obsessive Fortnite habits and how it contributes towards Fortnite's growth

Option 3

Article title

Zoom Video Communications

Article Evaluation

Relates to Civic Tech in the sense that society's lives have moved towards online platforms for communication such as Zoom. What are the implications of this transition and shift on how society perceives the government and or activism in the form of an online presence. Article information could be updated. Since Zoom is constantly being used by millions of users each day, there must be news that can be updated towards its Wikipedia page, since the last source/citation was from July 2020.

Is the article's content relevant to the topic? Yes Is it written neutrally? Yes Does each claim have a citation? Yes Are the citations reliable? Yes Does the article tackle one of Wikipedia's equity gaps? No

Sources

https://finance.yahoo.com/news/zoom-updating-messaging-platform-could-234511420.html <-- Insight on Zoom's messaging upgrade & investment

https://www.tomsguide.com/news/zoom-2fa-setup <-- Highlights Zoom's security updates

Option 4

Article title

Low Information Voter

Article Evaluation

WikiProject Politics; Start-Class. Seems outdated and oversimplified/lacking adequate information. No visuals/graphs/pictures/media.

Is the article's content relevant to the topic? Yes Is it written neutrally? Yes Does each claim have a citation? Yes Are the citations reliable? Yes Does the article tackle one of Wikipedia's equity gaps? No

Option 5

Article title

Open-access poll

Article Evaluation

WikiProject Internet; Start-Class. Outdated and oversimplified/lacking adequate information. No visuals/graphs/pictures/media.

Is the article's content relevant to the topic? Yes Is it written neutrally? Yes Does each claim have a citation? Yes Are the citations reliable? Yes Does the article tackle one of Wikipedia's equity gaps? No

Article evaluation

• Everything in the article is relevant to the article's topic. There is nothing that distracted me.
• The article is neutral (it discusses many regions, thus it provides a global perspective). There were no claims, or frames, that appear heavily biased toward a particular position.
• The information under Civic Tech in Asian may be underrepresented in related to other global regions. However, this may be an accurate representation of the lack of Civic Technology presence in Asia. Other than this, the viewpoints are not overrepresented or underrepresented.
• The citation links work, and the sources support the claims in the article.
• Yes, each fact is reference with appropriate, reliable references. The information comes from reputable websites such as the country's official Government website, public & official company statements, and reputable global non-profit organizations. There are neutral sources and are not biased.
• No, the information is not out of date, not do I think there is something missing that could be added. There are credible sources that are as recent as July 2020 that is relevant to the article's topic.
• The conversations on this article's talk page mainly consist of fact checking and contributions from research or academic institutions.
• The article is a part of several WikiProjects. The projects and its ratings are as follows:
  • WikiProject Technology (Rated Start-class)
  • WikiProject Politics (Rated Start-class, Low-importance)
  • WikiProject Sociology (Rated Start-class, Low-importance)
  • WikiProject Internet (Rated Start-class, Low-importance)
  • WikiProject Freedom of speech (Rated Start-class, Low-importance)
  • WikiProject Globalization (Rated Start-class, Low-importance)
• The way Wikipedia discusses this topic differs from the way we have talked about it in class in the sense that Wikipedia offers a more objective, factual account of the topic, as opposed to the discussion based approach we utilize in class, where students are able to offer their insight, opinions, and observations to advance the topic.

Article Topic: Civic technology

Evaluation Link: Talk:Civic technology#Reliable References and Sources