U.S. intelligence: Russians used AI to create faked video smearing VP nominee Tim Walz with abuse allegations. It includes “testimony” by person with same name as a student, but is not the student.

​

They also aim to “instigate physical violence in the fraught period after voters cast their ballots” by stoking “threats towards poll workers, amplifying protests & encouraging protests to be violent”

Americans in Russia creating disinformation: “In a two-day span 103 new sites appeared” that mimic news outlets.

How to spot a website generated by artificial intelligence

Some good tips but risks recreating the CRAAP test

https://observers.france24.com/en/how-to-spot-a-website-generated-by-artificial-intelligence and https://houstonpost.org/

​

Moscow’s Tools

​

​

​

​

• In North Carolina residents shared false information that a dam was about to burst, prompting hundreds of people to evacuate and diverting first responders. �
• McNeill, 43, had a panic attack…she knew she didn’t have enough gasoline in her car to evacuate from yet another disaster. But it was all a hoax.“The second storm turned out to be a nonsensical rumor. And when you are with limited cell service and you come across misinformation, you can make a split-second decision that can have life-or-death consequences.”�
• A viral falsehood was a claim that government officials planned to seize the flooded town of Chimney Rock, N.C., and bulldoze bodies under the rubble [which] garnered millions of views. One user suggested “a militia to go against FEMA,” in a post that had received more than half a million views.�
• A debunked conspiracy theory circulated about FEMA spending disaster relief money on helping migrants who are in the country illegally.

​

Fact checking organizations debunked the claims but they go viral and new ones emerge every day

https://www.washingtonpost.com/politics/2024/10/04/trump-hurricane-helene-fema-money-misinformation/

https://www.washingtonpost.com/politics/2024/10/04/no-biden-didnt-take-fema-relief-money-use-migrants-trump-did/

States have had to create sites tracking and debunking misinformation. North Carolina’s:

FEMA’s “Rumor Response” site

https://x.com/PatriotLiv/status/1843008483995918371

The margins and mainstream merge

​

​

• Musk, whose 200 million X followers make him the platform’s most influential user, sent 12 posts Friday…based on false or misleading claims. He labeled FEMA’s actions “treason” and endorsed a post saying VP Kamala Harris and Homeland Security Secretary Alejandro Mayorkas “should be jailed for life for spending FEMA money on illegal aliens and not Americans.”�
• Mark Robinson, Republican nominee for governor of North Carolina, has repeatedly alleged there has been no state government response to Helene, even though Gov. Roy Cooper (D) has traversed the affected areas, visiting emergency operations centers and meeting with storm survivors.��
• Trump has amplified the false claim about migrants [taking FEMA money frm Hurricane victims] during campaign rallies and on his platform, Truth Social.

Conspiracy theories and misinfo weaponized & spread by politicians using rhetorical techniques described by DiResta

“Unacceptable!” ��“Can you believe this?” ��“This is what they do”�

“We can’t let them do this”

�“Big if true,” etc.

​

�Project idea - follow an account like Gaetz’s (or Musk’s), look at use of conspiracy theories, how fringe/extremist accounts retweeted, language used, connections, how extremist themes and memes circulated, etc.

​

Or use DiResta as lens to explore how influence, propaganda & persuasion change online as the barriers between mainstream politicians, activists, influencers, merge

https://x.com/mattgaetz/status/1842778042659131469�https://x.com/mattgaetz�https://x.com/ImMeme0/status/1843054494193709148

The account Gaetz is amplifying

​

Genre bending, textual blending, remix

• Take an article from the Daily Aztec and rewrite it in the style of Gossip Girl. �
• You are a master chef. Read the U.S. Declaration of Independence like a recipe for a new country.�
• You are a great lover. A veritable ChatJuan. Read the following scientific article like a love letter. �
• Take the monologue from Fight Club and rewrite as if channeled by Elmo from Sesame Street.

Exploring creative applications (payful prompting)

1. Write an operatic libretto based on the recent presidential debate between Trump and Harris. Take the results, then transform it into a rap battle with Trump channeling Big Sean, and Harris channeling Kendrick.��[AI was annoyingly coquettish. Refused to “go there.” I told it the results were pathetic, anodyne, and uncreative, and challenged it to come up with a creative story idea.]�
2. Result: Steampunk Dinosaur Detective Noir in a Haunted Submarine�"A private eye velociraptor with a top hat and steam-powered monocle is hired to solve a murder mystery aboard a haunted Victorian-era submarine. As the dinosaur detective navigates through ghostly apparitions and mechanical contraptions, he uncovers a conspiracy involving a cursed treasure, time travel, and the ghost of the submarine's mad inventor.”�
3. Sample creative explorations by instructor

Jevonia Harris and Lauren Kelley, “From Algorithms to Insights: Understanding AI's Potential in Education”

​

COMBINING

A dynamic watercolor surreal painting of a beautiful Black woman with an Afro hairstyle. Her Afro is entirely made up of rhonocolus flowers, with deep purple hues, arranged to depict the Fibonacci series. The flowers seamlessly form the shape and flow of her hair, capturing the elegant spiral and harmonious design. The painting uses vibrant and fluid watercolor techniques to blend her features with the floral Afro, emphasizing the Fibonacci pattern.

​

AI has the Creative Power to Unite Experience and Learning

Image: DALL-E

“Jagged frontier”: Image editing, revision, and graphic design

​

“Two years ago, the Biden administration issued� a blueprint for an AI Bill of Rights aiming to address

“unsafe or ineffective systems,” “algorithmic

discrimination,” and “abusive data practices,”

among other things. Then, last year, Biden built

on that document when he issued his executive

order on AI…. Several weeks later, the UK hosted an� international AI Safety Summit that led to the� serious-sounding “Bletchley Declaration,” which� urged international cooperation on AI regulation. The� risks of AI fakery in elections have not sneaked up� on anybody.

​

Yet none of this has resulted in changes that would

resolve the use of AI in U.S. political campaigns. Even

worse, the two federal agencies with a chance to do

something about it have punted the ball, very likely

until after the election.”

​

Tracking and analyzing some of the AI-based�persuasive strategies used in the 2024 elections.

​

Patterns, tactics, strategies, changes, potential effects?

Final Paper/Project Ideas

Election-related papers and projects. New material available.

​

1. Searchable collections of social media activity figures and organizations�
2. Speech Transcripts�
3. Reports and repositories of data on election interference

​

Reports and repositories of data on election interference

AI-generated transcripts and data repositories of political figures/influencers

https://fight.fudgie.org/search/show/aj/

https://trumpstruth.org/

(The Bulwark was founded by conversative �writers Bill Kristol, Charlie Sykes, & Sarah Longwell who are critics of Trump)�

To narrow scope - most shared/retruthed posts: thematic analysis, frame analysis, persuasive strategies, etc. Shared material seems to be more extreme (distance). Patriot4Life currently #1

Musk is leveraging X’s power as a source of real-time news for millions of users to try to sway the outcome of the 2024 presidential.

​

Musk has already asserted his voice in the political sphere for months. He’s pushed racist conspiracy theories, obsessed over the “woke mind virus”… and warned of the country’s impending “doom” if a “red wave” does not materialize in November. But now Musk, who is the most followed user on X, has shifted to more direct campaigning for former President Donald Trump

​

Elon Musk is using his social media platform to promote misinformation about the presidential candidates in the lead up to the November election, amplifying false claims Wednesday about a Trump rally bomb threat and immigrants eating pets in Ohio….his posts regularly finding their way atop users’ feeds.

Five secretaries of state sent a letter to Elon Musk Monday imploring him to fix X's AI chatbot after it shared misinformation about the 2024 presidential election.

Suno.ai 

• REDDITERS: “Okay. I officially cannot tell WTF is AI generated music anymore. Wow.”�
• “Literally just give it a premise and a genre and boom, you're done. Sounds dumb but my partner and I use it to compose stupid little songs to each other about how our day was.”

​

• “You can feed it original music and lyrics. It's getting really good at it. Mine has been trained from my recording, so I hear variations of previous songs in new ones that were created later.”

Suno.ai 

• Similar to uploading written texts and using them as GPT base. �
• Can give it your own lyrics or edit ones it creates. �
• Can use out of copyright songs and remix, perhaps making possible the kind of work Moby is famous for. For example, taking old spirituals and folk songs and combining with house music.

Suno.ai 

Torturing my kids: https://suno.com/song/3c886cab-acbb-4541-9062-c060666a3d15

​

Can remix existing open-source songs: “Covers of the song Once” by @io https://suno.com/playlist/e18572ff-7693-49f4-82ca-c2c93b9b2ac2

​

“Suno Star” remix by songwriter https://suno.com/song/8eac4ad1-e402-4715-b174-f7c7002a72f2

​

“A playlist of covers i made after recording myself singing in the shower” https://suno.com/playlist/7adc2d14-69d5-46da-99df-71a08974d877

• AI companies use our social media content to train LLMs.
• Social, cultural, communicative and community knowledges extracted and monetized.
• Companies claim can take/profit from our content based on “fair use,” a principle they’ve opposed in every other context.

​

​

​

​

Our social, cultural & communicative knowledges extracted and monetized.

Your thoughts, writing, stories, family photos, etc. fuel LLMs.

​

​

​

​

​

Your thoughts, writing, stories, family photos, etc. fuel LLMs.

​

Human “batteries.” The Matrix reinterpreted?

​

​

​

​

​

​

​

​

​

​

Bias continued

1. In the past couple days, I took a Kaggle mini-course on AI ethics that I thought would be worth sharing to the group (no coding necessary). I found it valuable and relevant. In one demonstration, the course shows how a toxic text classifier algorithm done inattentively will flag targeted minority identity labels as toxic (e.g., Black, Muslim) just by virtue of being associated so often with hate speech in online training sets. (More on this issue is also detailed in a Kaggle competition referenced by the course.) Intro to AI Ethics. Instructors: Alexis Cook (Kaggle), Var Shankar (Responsible AI Institute)

​

Professions project

1. Last week I recorded and published an episode of the Social Work Podcast where I interviewed GPT4o. In social work we distinguish expertise through experience, rather than education or training. My podcast includes interviews with people whose expertise runs the gamut. I've been interested in doing an episode on AI for quite a while, but couldn't land on the right angle. After GPT4o was released I realized I wanted to interview AI as an expert on AI. The conversation was spontaneous - I didn't preplan the questions and I didn't edit the answers. While the discussion isn't about education, there's lots of ways instructors could interrogate GPT4o's responses, the future of content creation, and the value of AI as an "expert" on AI.
1. Singer, J. B. (Producer/Host). (2024, May 22). #137 - The Role of AI in Child Welfare and Community Organizing: Interview with ChatGPT-4o [Audio Podcast]. Social Work Podcast. Retrieved from https://www.socialworkpodcast.com/2024/05/ChatGPT-4o.html�COULD SHOW STUDENTS AND SUGGEST THEY RUN A SIMILAR EXPERIMENT. But analysis of the output is key. Conversation on list:  https://groups.google.com/d/msgid/ai-in-education/4a31852e-6cf2-4dd8-9fd5-b626d6e49479n%40googlegroups.com.

​

• An AI chatbot helped Americans who believe in conspiracy theories “exit the rabbit hole”
• The uncle who believes 9/11 never happened. The next-door neighbor who thinks Biden stole the 2020 election. The Nieman Lab editor who’s been known to wonder if aliens do exist and the U.S. government is covering them up.
• You probably don’t want to talk to these people and convince them that they are wrong. But what if an AI chatbot could do it for you? That’s exactly what a group of researchers just did. In their peer-reviewed article “Durably reducing conspiracy beliefs through dialogues with AI,” featured on the cover of Science this week, Thomas Costello of American University, Gordon Pennycook of Cornell, and David Rand of MIT explain how they put 2,190 conspiracy-believing Americans in brief but detailed conversation with the large language model GPT-4 Turbo.
• https://www.niemanlab.org/2024/09/an-ai-chatbot-helped-americans-who-believe-in-conspiracy-theories-exit-the-rabbit-hole/

Microdosing– search google TRENDS. 1000% increase over last 10 years. July 3, 2024. NPR. Use as part of lecture notes on search literacy. Can examine by state, city, etc. General search on microdosing, search on “microdosing mushrooms” (amazing).�

Sabo said he was motivated to post the signs because of reports that a Venezuelan gang forced the closure of an apartment complex in Aurora. The City of Aurora and Aurora Police have said the closure was not due to gang activity but rather years of neglect by a landlord now facing charges for neglect of the property.

​

In a YouTube livestream on Wednesday, prior to the posting of the signs on Colfax, Sabo complained about Venezuelans and other Latinos, and used an ethnic slur for the Latino journalist who covered the racist signs posted outside the Capitol and the closure of the apartment complex in Aurora. He listed specific Latino nationalities he would use as “target practice” during a conflict

​

“If a civil war breaks out and the f---ing Guatemalans and Ecuadorians are like coming over the fence and stuff like that, it is straight up target practice,” Sabo said.

​

​

https://www.9news.com/article/news/local/next/next-with-kyle-clark/credit-racist-signs-denver-bus-stop/73-4983ef41-702f-486d-92af-14cf3a70214a

• An Instagram account linked to Sabo — a right-wing street artist known for controversial art criticizing progressive policies and candidates — posted photos of the three Denver signs and a fourth at the intersection of Nome Street and East Colfax Avenue in Aurora.
• That intersection is near an Aurora apartment complex recently shut down for city health and building code violations, uprooting hundreds of Venezuelan migrants.
• https://www.denverpost.com/2024/08/29/kamala-harris-bus-stop-signs-denver-rtd-police-immigration/

https://www.cpr.org/2024/09/06/venezuelan-gang-claims-aurora-republican-campaign-event-fabbricatore-boebert-crow/

• National headlines: armed Venezuelan gang taking over an apartment building in Colorado.
• Based on cryptic video circulated on social media then TV.
• The video showed men with guns in an Aurora apartment complex, allegedly filmed by a scared resident. Nobody can confirm. Residents say its untrue.
• Trump: “thugs from Venezuela are taking over big areas, including buildings. They have their big rifles.”
• Colorado's Republican Party sent a fundraising letter claiming the state is under violent attack, and Venezuelan gangs have taken over Aurora.
• But Aurora's mayor, Republican Mike Coffman, denies there's a gang takeover. He calls the owners out-of-state slum lords, and critics argue the owners are alleging a gang takeover to shirk their responsibilities.

​

​

https://www.npr.org/2024/09/06/nx-s1-5100765/aurora-police-chief-says-theres-no-evidence-that-venezuelan-gang-took-over-apartment

Building residents called a press conference saying, “criminals don't live there at all.”

​

But tenants walked reporters through apartments infested with rodents, bed bugs and broken appliances. They say the company that manages them, CBZ Management, doesn't return their calls for repairs. In early August, CBZ emailed reporters a statement from a public relations firm, alleging a, quote, "violent takeover" by Tren de Aragua of multiple buildings the company owns.

​