Exploration #105

Hate Longscrolling? Jump Around!

• If You Click Only One…

• Things to Think About…

• Things to Know About…

• And finally…

Hi all. I’ll be speaking tomorrow (Thursday) at PMVG’s Public Media TechConnect ‘24, and then hosting a roundtable at breakfast Friday morning. If you’re there, come up to say “hi” and hang out for a bit.

Here in the newsletter, we’re talking about Suno’s generative music, public trust in ChatGPT, Sora’s first music video, working in the Apple Vision Pro, and the potential impact of AI on SEO. But first…

April’s 3rd Thursday Webinar

A reminder that next Thursday, April 18, the Public Media Innovators PLC is hosting its 3rd Thursday Webinar. In PBS KIDS: Innovating Accessibility in Children's Media, Jen Rodriguez, Director of Research and Inclusive Design; Nita Mandar, Director of Product for the PBS KIDS Games App; and Kelly Corrado, Director of Game Tooling & Analytics, will share their latest advancements in accessibility within children's media. You’ll get a firsthand look at, as well as a demo of, the approaches and tools that PBS KIDS is employing to ensure their content is inclusive and accessible to all audiences. You can register for the webinar here.

Must Be Something In the Water

Next, some colleagues here at the University of Nebraska-Lincoln’s Center for Emerging Media Arts are developing a public-private partnership model that might be right for helping some of you prototype ideas and dip your toe into the emerging media pool. The initiative is called Edgeworks and you can learn more about what they are bringing to the table via their website.

AI Music: 3 Chords & Some Truthiness 

A couple weeks back I referenced the generative music app Suno. I finally had a chance to play around with it more this week, and I was genuinely impressed. Not that it was creating instant classics, but it really is creating songs that were nearly fully baked. I asked it to create a pop punk anthem about this week’s Public Media Tech Connect in Las Vegas, and it gave me back “Electric Dreams” (click the title to listen). Are the lyrics inane? Yes. Are they nonsensical? No. And is it a pop punk song? Yes, even including a weird genre flirtation with dubstep in the breakdown. That song was generated from a zero-shot prompt. One and done. No coaching. Emilia David reported in The Verge that AI-generated songs are getting longer, not necessarily better. But I’m not sure I agree with that take.

Regardless, this has the recording industry concerned, resulting in Billie Eilish, Pearl Jam, Nicki Minaj Among 200 Artists Calling for Responsible AI Music Practices. You can read their open letter here. I appreciate that it acknowledges: “Make no mistake: we believe that, when used responsibly, Al has enormous potential to advance human creativity and in a manner that enables the development and growth of new and exciting experiences for music fans everywhere.” It goes on to say, “Unchecked, Al will set in motion a race to the bottom that will degrade the value of our work and prevent us from being fairly compensated for it.”

I remember when music was scarce. When you had to make the trek down to the record shop by campus that smelled funny to exchange lawn mowing money for 12” mixes from the UK. And then pooling scarce resources with friends via mixtapes. Unless you could find it used (which itself was legally disputed at one point), a single piece of physical media could cost you $15.99, close in price to a monthly music subscription today that lets you access nearly the entire history of music.

Thankfully, those days are gone. On the supply side, computers have made creating and distributing music increasingly efficient and artists adapted. In the late 1970s and early 1980s, young artists like Gary Numan (“Cars”) and Orchestral Maneuvers in the Dark (“If You Leave”) famously walked away from guitars because the big sounds they wanted to make came from computers. Sampling, whether derived from the turntable and hip-hop culture of the 70s or from the ability to quickly record and sample real world sounds, opened up even more creative avenues for artists. Generative audio feels like the next frontier to be artistically integrated into the musician’s creative toolkit.

On the demand side, music is deeply personal. Some songs grow with you, like good literature. Some are personal time machines. Some make you question civilization. Overplayed hits become cringe and then become guilty pleasures (listening to you, Yacht Rock). To date, all that has tied back, for beter or ill, to a human touch. And while instrumental music, especially soundtracks, can be deeply moving, emotional connections to songs often come down to words.

If I had a business making needle-drop music, I’d be looking to pivot. While the poetry may not be there yet, listeners will connect with the musical level of AI songs and soundtracks. Public media needs to start thinking about the balance between ethical and economic sense.

Okay, on to the links…

If You Click Only One…

It’s 2024 - How will Public Media Innovate? - (Mikey Centrella - PBS Hub) - This is the inaugural post in the PBS SPI Team's "Thinking Beyond Screens" blog series. Friend-of-the-newsletter, Mikey Centrella lays out his take on public media innovation in 2024 with 5 trends. I’m especially down with trend #4.

Things to Think About…

Meta’s AI image generator can’t imagine an Asian man with a white woman (Mia Sato - The Verge) - This is ironic given Zuckerberg's wife is Asian-American, which also leads me to believe this will be fixed tout suite. And while I don't think of Meta's image generator as being an industry leader, Sato makes good points about Asian depictions in media that are worth a read.

Digital humans: the relatable face of artificial intelligence? (Evan John - CNN) - A couple weeks back I included a cover image to the newsletter that was a digital human...an AI models idea of what a PBS viewer (and NPR listener) looked like. My South-by summary edition had a room full of them. Of course these are just stills, but others are taking this a step further with interactivity.

Do AI Systems Deserve Rights? (Eric Schwitzgebel - Time) - This won't be the last time you hear this question. Though only a small number of individuals are, at this point, trying to argue that AI tools like ChatGPT, Gemini and Claude, are sentient, as the models get better more and more people will start arguing this point.
—This session from SXSW 2023 was one of the most interesting of that year’s conference: A Practical Guide for Whether Your AI Is Sentient
—See also: "And Finally..." below.

Things to Know About…

Americans’ use of ChatGPT is ticking up, but few trust its election information (Colleen McClain - Pew Research Center) - We're seeing similar reactions specifically in Nebraska (more on that in a future exploration). To those Americans I say, bless all of you. While I'd rather we had AI tools that earned our trust, that's not where we are in the development of this technology and I'm glad Americans developing media literacy around it.

Washington state judge blocks use of AI-enhanced video as evidence in possible first-of-its-kind ruling (Tom Stelloh - NBC News) - Here's an interesting development that journalists especially should note.

OpenAI's Sora just made its first music video and it's like a psychedelic trip (Cesar Cadenas - TechRadar) - Remember two things. First, AI video is only going to get better from here. Second, the early days of human-produced videos in the 1980s were also pretty trippy. At least today we have the technological limitations to blame. Exhibit A.

How the Ad Industry Is Making AI Images Look Less Like AI (Katie Deighton - Wall Street Journal) - A good piece showing the value of experimentation and transparency in the results.

Will traffic from search engines fall 25% by 2026? (Danny Goodwin - Search Engine Land) - Rumors of search's demise have been greatly exaggerated before, even by Gartner (the source of this article). But when you start to consider the advances made by Perplexity and its kind, and the rise of TikTok as a legit reference source for younger audiences, it starts to seem less far-fetched. If you're organization has an SEO strategy (and your organization really should have an SEO strategy), you need to be considering the potential impact of AI on SEO.
—But, as Barry Schwartz report in Search Engine Land, it's not like Google will go quietly: Google starts testing AI overviews from SGE in main Google search interface
—And, as Chris Vallace reports in the BBC: Google using AI to come up with search answers in UK trial 
—Of course, this begs the question, will AI optimization become the next SEO? Mark Sullivan of Fast Company address that via his LinkedIn newsletter: Brands are growing more concerned about how they are perceived by ChatGPT

AI-powered search engine Perplexity launches an AI-generated daily news podcast (Andy Meek - Boy Genius Report) - Like most things AI, it only feels slightly off. The pauses between sentences will occasionally make you wonder who edited it. And the voice is good be occasionally lacks the intonation you'd expect, making you wish the reader would get some training.
—Take a listen, it's better than you think it's going to be: Perplexity.ai on YouTube
—Perplexity may also be trying to supplement is subscription reveue with ad revenue, as reported by Trishla Ostlaw in AdWeek: Gen-AI Search Engine Perplexity Has a Plan to Sell Ads

Welcome to the AI gadget era (David Pierce - The Verge) - I recently moved, which meant confronting all the e-waste from a similar gadget phase that Web 2.0 brought the world a decade ago. So, shocker, I'm kind of cooling on the idea AI gadgets right now, but it's still a space to watch.

First VR developer integrates with OpenAI, setting stage for no-code VR development (Zoltan Vardai - CoinTelegraph) - It seems logical that AI is going to create a sensation, once that chocolate and peanut butter fully integrate. But keep in mind that variations on tools like this will also be integrated into platforms like Roblox and Fortnite as well.

I worked exclusively in Vision Pro for a week—here’s how it went (Samuel Axon - Ars Technica) - I've been on some meetings with a couple of Personas (kudos to our iOS dev Rob and PBS's Nick Licintra for being pioneers there) and, yeah, they're still a little 'uncanny valley.' That aside, I continue to feel like head-mounted displays (HMDs) like the Vision Pro will be productivity tools before they are personal entertainment tools.
—In other AVP news: David Heaney reports in UpLoadVR that Apple Vision Pro Now Has 3D 'Spatial Personas' For Virtual Hangouts & Meetings  

LG Launches Business Metaverse (Rory Greener - XR Today) - H/T to Tom Ffiske for this one. This seems smart to me. The metaverse is going to be niche and work-based before it becomes prolific in homes. If this is successful, I'd expect you'll see a lot more of these across a variety of industries. In fact, I'm a little surprised, no one has created this type of B2B community space for broadcast tech. If anyone knows of any, DM me.

Spotify adds video learning courses in latest experiment (Jon Porter - The Verge) - will Spotify be a part of your video distribution strategy in 2025? Keep an eye on this experiment.

Using Premiere Pro’s latest AI audio tools (Shiv Rajagopal - RedShark News) - For those using Premiere Pro in their shops (or managing those who use it), you may find this useful.

NYT to soon offer most articles via automated voice (Sara Fischer - Axios) - As Fischer points out, "The move will help position the Times as one of the biggest audio news companies globally." Given that a not-insignificant portion of our industry is high quality audio news, this feels significant. Of course, the Economist and others have been doing this for year, so maybe this is solely a win for accessibility, a not the threat I'm imagining it to be.

And finally…

Have a creative, productive week!