Exploration #95

Need a Shortcut? Jump Then Fall

• If You Click Only One…

• Things to Think About…

• Things to Know About…

• And finally…

Hi all. Before we dive into the news and thoughts on emerging media from the past week, we need to pay some serious attention to what internet trolls are doing to Taylor Swift.

Here’s the recap: A little over a week ago graphic deepfake images of Swift started showing up on Twitter/X (there is competing reporting on where the outbreak began). According to Jess Weatherbed in The Verge: "One of the most prominent examples on X attracted more than 45 million views, 24,000 reposts, and hundreds of thousands of likes and bookmarks before the verified user who shared the images had their account suspended for violating platform policy. The post was live on the platform for around 17 hours prior to its removal."

Then Swifties, the Taylor Swift fan army which contributed so much to the 2023 US economy that even the Fed noticed, did what only they could do: flooded Twitter/X with real concert footage and images in an effort to swamp deepfake content with legit content. The response from Twitter/X has been to temporarily block searches for “Taylor Swift,” which only serves to punish the victim. Meanwhile there are renewed calls for legislation from the White House and Congress. You might be fortunate enough to live in a state that is regulating AI, and maybe even one that has laws against sexually explicit nonconsensual deepfakes. But at the national level it all essentially amounts to “thoughts and prayers.”

As Vice trenchantly put it: Taylor Swift Is Living Every Woman’s AI Porn Nightmare. Thanks to gAI tools, these type of images have become incredibly easy to make. And while, prior to Swift being targeted, public attention to this issue focused on politicians and a couple of actors (read: men), the preponderance of deepfake imagery — and nearly all nonconsensual pornographic imagery — targets women (AFAB or self-identified).

This is not the first time we’ve discussed this issue here. OG readers may recall that last fall Matt Burgess illuminated the scale of the problem via Wired in Deepfake Porn Is Out of Control: “At least 244,625 videos have been uploaded to the top 35 websites set up either exclusively or partially to host deepfake porn videos in the past seven years, according to the researcher, who requested anonymity to avoid being targeted online. Over the first nine months of this year, 113,000 videos were uploaded to the websites—a 54 percent increase on the 73,000 videos uploaded in all of 2022. By the end of this year, the analysis forecasts, more videos will have been produced in 2023 than the total number of every other year combined.”

Swift's role as a next-gen Oprah(x10) has raised awareness of this issue, but thousands of women and teens have already been victims of deepfakes and there is no federal path to justice. Guys, take note, because women already get this: this type of content is violence against women.

Now, Public Media can’t fix this alone. If the only immediate solution for Taylor Swift on Twitter/X is to make her content unfindable, then the system is deeply flawed. So what can we do to be part of a solution? First, understand the larger contexts at play here. This isn’t new. It’s been an online issue since the early days of Photoshop and networked computers. AI has simply accelerated that harm (as it has accelerated so much else in tech). And that harm is reflective of a larger societal issue of violence against women — as well as anyone who doesn’t identify as a white cisgendered male. Technology, as it often does, is amplifying the human experience, and that can be traumatic, especially for women.

Next, tell and elevate more of these stories…especially the local ones. Public awareness (beyond Swift) is key, because this public moment will fade, but the problem will not. This means telling stories about the local legislative environment that supports or inhibits recourse for victims as well.

Finally, take what you learn and be a thought leader. The links in this column can be a resource. As someone who is interested in the cutting edge of communication technology, influence the world around you by making it more literate to these problems and the tech that generates them. And if you have a young man in your life, have the hard conversation and set the expectation for a responsible use of these tools. They will only get more powerful and more pervasive as they (the young men and the tools) get older.

Okay, on to the links…

If You Click Only One…

Taylor Swift Deepfakes Show What’s Coming Next In Gender and Tech – And Advocates Should Be Concerned (Ariana Aboulafia & Belle Tore - Tech Policy Press) - I’ll let this piece, from two former legal fellows to the Cyber Civil Rights Initiative, speak for itself.

Things to Think About…

How Creators Can Use AI To Be More Efficient And Still Be Creative (Jennifer Kite-Powell - Forbes) - Up to this point, this exploration has dwelled on the use of gAI tools to make offensive content. But as we normally see in this newsletter, these tools can be used for good too. This paragraph, from late in the piece, resonated with me: "However, one thing is sure, just as Smartphones have empowered everyone to take better pictures, AI tools will lower the barrier for those making videos, leading to a flood of new films available online," said Armitage. "Those who have good stories but have lacked the skills and budgets to tell them might finally be able to bring them to life." A future like that can only benefit the mission of public media.

Most Top News Sites Block AI Bots. Right-Wing Media Welcomes Them (Kate Knobs - Wired) - Does your station allow AI Bots to scrape your web content for model training? In Nebraska, we are still allowing those bots to crawl our site because, as any good editor knows, garbage in = garbage out. Our content is free to the public, and I argue that making it available to these bots is an extension of our mandate for universal service. But I get why those media organizations with profit motives want to license their content to these AI companies (through cash, barter or legal arbitration). Either way, if you haven’t made a decision on this, well, you’ve still made a decision on this. And it's likely your organization’s content is being scraped by these bots.

3D scanning: we recreated a sacred South African site in a way that captures its spirit (Stephen Weasels, Benjamin Schoville, Jayne Wilkins, Sechaba Maape - The Conversation) - Documenting local places of cultural significance and putting them in historical context should be a big part of public media's mission in the 21st century. Already we are seeing the first versions of these through the virtual tours created by Wisconsin, South Carolina and Georgia. Our education department has created a couple in Nebraska as well. This piece from South Africa advocates for going beyond the documentation of space and including elements that evoke story to create a sense of place; something at which we already excel.

Things to Know About…

Researchers Say the Deepfake Biden Robocall Was Likely Made With Tools From AI Startup ElevenLabs (Kate Knibbs - Wired) - We touched on this last week, but in light of the Taylor Swift story, it's worth checking in on this next chapter. I’ve mentioned ElevenLabs audio tool here before, as we're testing it for legit audio work in Nebraska.
—Though, as Miles Klee reported almost a year ago in Rolling Stone, bad actors have been practicing for this moment: Fake Biden Speeches Are the Hottest Trend in AI Voice Tech
—And, as Bryson Masse reported last fall in VentureBeat, it's not just Joe: Confessions of an AI deepfake propagandist: using ElevenLabs to clone Jill Biden’s voice
—For some historical context, James Vincent reported a year ago in The Verge on the popularity of ElevenLabs amongst bad actors: 4chan users embrace AI voice clone tool to generate celebrity hatespeech

The Vatican’s top expert on AI ethics is a friar from a medieval Franciscan order (Frances D'Emilio - AP) - Again, trying to balance out the dark with a little light. Here's a key quote from Friar Benanti: “It is a problem not of using (AI) but it is a problem of governance,’' the friar said. ”And here is where ethics come in — finding the right level of use inside a social context.” One of the things I noted here what that his optimism is directed toward use of AI for science, but not so much for communication.
—Related, by Benj Edwards in Ars Technica: AI-generated puffy pontiff image inspires new warning from Pope Francis
—Friar Benanti, in the profile, references the extraction of resources from the global south (“digital colonialism,” as others have called it), this recent piece in The Guardian by Josh Taylor illuminates why those are concerns: Precarious conditions of AI ‘ghost workers’ revealed by Google termination of Appen contract, union says

OpenAI updates ChatGPT-4 model with potential fix for AI “laziness” problem (Benj Edwards - Ars Technica) - At the end of last year, we talked about how ChatGPT seemed to be slacking off for a winter break. I know I personally noticed this in both ChatGPT Plus and DALL-E 3. This past week, OpenAI seems to have addressed this problem, but with no explanation as to why it happened in the first place.

Why publishers are hesitant to add their chatbots to OpenAI’s GPT Store (Sara Guaglione - Digiday) - Ah, the sound of hype deflating.

I literally spoke with Nvidia’s AI-powered video game NPCs (Sean Hollister - The Verge) - We'll be digging more into games in the coming weeks, but this nexus of AI and gaming gives you a glimpse of the future.

Want Better AI Images? Ask a Chatbot! (Daniel Nest - Why Try AI?) - For over a year now, I've been a fan of the combinatorial powers that come from using multiple AI tools. Nest's tutorial is a good way of using a chatbot to help you craft a generative art prompt that gets you closer to the image in your mind's eye.

AI poisoning tool Nightshade received 250,000 downloads in 5 days: ‘beyond anything we imagined’ (Carl Franzen - VentureBeat) - Last week I mentioned that Nightshade had been released. Turns out, it has proven to be quite popular.

Google’s latest AI video generator can render cute animals in implausible situations (Benj Edwards - Ars Technica) - Generative video is coming. It's not *there* yet for professional use, but announcements like this show the potential.

New education features to help teachers save time and support students (Shantanu Sinha - Google for Education) - We've talked recently about Microsoft beginning to gear its gAI offerings toward educators. Not surprisingly, Google is doing the same.

Microsoft Teams now supports 3D and VR meetings (Tom Warren - The Verge) - Finally! This has been 'just around the corner' for the better part of three years now. Early reviews are mixed, and this timing is odd given that Microsoft culled its on-staff xR herd in 2023. But I’m still looking forward to trying it out.
—Here is Microsoft's announcement: Bring virtual connections to life with Microsoft Mesh, now generally available in Microsoft Teams

Developer interest in the Apple Vision Pro is positive, not negative (Tom Ffiske - Immersive Wire) - As the straight-down-the-middle headline indicates, Ffiske offers a hype-resistant assessment of the Apple Vision Pro a week after orders started and a week before units start to arrive on developer's doorsteps. I’m sure we’ll start to hear more reviews next week after the public gets its first at-home taste of the ‘face computer’ this weekend.

XR Association Releases "State of the Industry Report," Offering Reflections on 2023 and a Look Ahead to 2024 (XR Association) - The source is a trade association, so you gotta take this with a grain of salt. But if you understand the source and its agenda, I think you can get a sense of where we are in the XR winter.
—Download the PDF here.

And finally…

Guns N’ Roses share AI-generated video for ‘The General’ (Eli Ordonez - NME) - There's a “Use Your Illusion” joke in here somewhere. Nevertheless, once you see the morphing effect of gAI video you'll start seeing it in a lot of places. The song is a "meh" (IMHO), but hey, it's a B-side.

Have a creative, productive week!