Exploration #97

Hate Longscrolling? Jump Around!

• If You Click Only One…

• Things to Think About…

• Things to Know About…

• And finally…

Hi all. Happy Valentine’s Day (if you observe)! This week, we’ve got some fresh AI news, more experience with the Apple Vision Pro (AVP), and a big deal in the world of gaming. But first…

A reminder that our second “3rd Thursday” webinar is February 15 (tomorrow, if you’re reading this the day it drops) at 1pET and will feature educator and generative AI power user Ethan Mollick. His newsletter, One Useful Thing, is a regular staple of the links shared here, and today is no exception. If you haven’t yet, there’s still time to register!

Twinning with Gemini

As anticipated in last week’s Exploration, Google did roll-out a rebranding of its Bard gAI tool, along with a new premium subscription tier and access to their most powerful Gemini model. Gemini’s screenside manner is definitely closer to Claude than ChatGPT. But the reviews are mixed. In Forbes, Charlie Fink commented that Gemini felt neutered by the copyright police. But some, including webinar guest Ethan Mollick (who had access to an early release of Gemini) really seemed pleasantly surprised by it.

In a week of intermittent testing, I actually think it’s regressed from Bard. Now, maybe that’s because the servers were getting slammed thanks to the new attention, but its ability to generate images was non-existent (it actually generated two images, then told me it didn’t have the capability to generate images, and then suggested I go try Bard instead), and a request to summarize large blocks of text also left it flummoxed. That said, it was a little better (if not perfect) at helping to edit this newsletter for typos.

While I’m sure, under the hood, it’s the competitor to ChatGPT that many say, maybe don’t give up that ChatGPT Plus subscription just yet.

You Down with AVP?

In Nebraska Public Media Labs, we have now had our Apple Vision Pro a little over a week. I asked our iOS developer, Rob Fenton, to jot down his reactions and he had this to say: “I keep thinking about a line from Contact, "They should have sent a poet." It's amazing. I've been giving a few demos a day since we received it and it's clear from the positive reactions everyone has had to Encounter Dinosaurs that immersive storytelling is going to be very powerful. It really feels like they nailed positioning of virtual objects in the physical world, and combined with the quality of the displays and the spatial audio, you really feel like you're part of the experience.”

And Amber Samdahl, from PBS Wisconsin, did the demo at her local Apple Store. Here's what she had to say when I asked for a hot take. “The Apple Vision Pro is a game-changer. In my experience, the stunning visual fidelity and seamless mixed reality of the AVP put it in a class far beyond any VR headset. Its elegant user experience, something Apple always champions, truly sets it apart from its competitors.All that said, with its current price point, the heavier weight, and the clunkiness of having an external battery, I will personally be waiting to buy one until the third version when Apple really gets its new devices right.”

For my part, I spent about 90 minutes in the AVP (and was lucky I had Rob guide me through the setup). The hand-tracking is, IMHO, better than anything commercially available. That, combined with the eye tracking, is what creates the feeling of magic you see referenced in reviews. It takes the intuitiveness of an iPad to an immersive, room-scale experience. And thanks to the high quality pass-through camera, which shows you your surroundings, it finds good balance between immersion and safety, quickly building user trust in the device. No walking into walls or be startled by someone walking into your ‘play space.’

It isn’t perfect, but it is as good a first version of a device as the iPad and iPhone were. You can truly see the potential. As both the price and the weight of the device come down over future generations, it’s easy to see how the AVP can become more integral to work and education. For those who live (or at least consume media) alone it could be a great media consumption device, but more attention needs to be paid to the social (headset-to-headset) aspects of usage for it to truly have a home in the living room. And as an tool for accessibility, it has undeniable potential to enhance quality of life.

If you can, I’d encourage you to schedule a demo at an Apple Store to get a feel for what the future holds.

Austin’s Power

We are four weeks out from the start of the 2024 South by Southwest Interactive conference in Austin, TX. If you are attending this year, please drop me an email: [email protected]. I'm pulling together all the attendees from public media into an email thread so we can share notes and, once on the ground in Austin, a group message thread organized by David Huppert from PBS NC.

If you haven't gone before, you should put it on your calendar for March 2025. Bar none, it's the best professional development conference for what we do as change agents in public media. For me, it's a week-long rumination about where media is headed in the next year, and it helps me set the strategic intention for Nebraska Public Media Labs as I look at our next fiscal year.

Via this newsletter and our March “3rd Thursday” webinar, you'll still get the inside scoop on what your public media colleagues are thinking about after their time at 'South-by' but there's really no substitution for the immersion in tech and culture you get there.

Okay, on to the links…

If You Click Only One…

Google's Gemini Advanced: Tasting Notes and Implications (Ethan Mollick - One Useful Thing) - Google gave Ethan early release access to their upgraded tool (another reason to register for the webinar with Ethan, Google has him on speed dial). So he's been able to conduct his brand of real-world use, and gives you a good sense of what you can expect as you put it through its paces for your work.

Things to Think About…

Artificial Intelligence in the News: How AI Retools, Rationalizes, and Reshapes Journalism and the Public Arena (Felix M. Simon - CJR) - There is a lot to crunch on here, and I suggest reading the conclusion first before digging into the meat of the report. I’ve pulled two nuggets from the Simon’s own summary that I thought worth elevating. First, "AI will play a transformative role in reshaping news work, from editorial to the business side. We are witnessing — to a degree — a further rationalization of news work through AI. It is important to recognize that the extent of this reshaping will be context- and task-dependent, and will also be influenced by institutional incentives and decisions." And then, "Productivity gains from AI in the news will not be straightforward. The benefits of AI to the news will be staggered. They will incur costs in the early stages and necessitate changes at the organizational and strategic level."
—Download the PDF directly from Columbia University here.

Artificial intelligence needs to be trained on culturally diverse datasets to avoid bias (Vered Schwartz - The Conversation) - Think pieces, like this one from an Associate Professor of Computer Science at the University of British Columbia, come around every few weeks. But the issue deserves that type of attention. If you watched our first “3rd Thursday” webinar, you heard Ernesto and Lauren discuss mitigating the biases baked into the tools. They were providing user-based mitigations. But a true solution would be to have less- or non-biased large language and multimodal models in the first place.

"AI native" Gen Zers are comfortable on the cutting edge (Jennifer A. Kingston - Axis) - I, for one, welcome our Gen-Z overlords. This piece very much jibes with what I see working with college students and recent grads as a part of Nebraska Public Media Labs.

Things to Know About…

Google’s Gemini is now in everything. Here’s how you can try it out. (Will Douglas Heaven - MIT Technology Review) - Here’s a good breakdown of how to access Gemini, including comparisons to ChatGPT Plus (its only real competition).
—Bernard Marr also put Gemini to the ChatGPT test in the Forbes piece, AI Showdown: ChatGPT Vs. Google's Gemini – Which Reigns Supreme?
—And here is Allison Johnson's take in The Verge: Google’s Gemini assistant is a fantastic and frustrating glimpse of the AI future

Labeling AI-Generated Images on Facebook, Instagram and Threads (Nick Clegg - Meta Newsroom) - Meta joins other major tech companies in agreeing to do their part to ID gAI images. But it's worth noting that, while the major generative art tools are aligning around this custom, open source generative art models will not have to comply with such guardrails.
—Related, from Emilia David at The Verge: OpenAI is adding new watermarks to DALL-E 3
—And Melissa Heikkilä provides some larger analysis in her intro column to the MIT newsletter, “The Algorithm": Why Big Tech’s watermarking plans are some welcome good news 

‘DignifAI:’ 4chan Is Editing Pictures to Clothe Women (Samantha Cole & Emanuel Maiberg - 404 Media) - Lest you think deepfake nudes were the only way guys could abuse women with generative art tools, we now have this.

Copilot gets a big redesign and a new way to edit your AI-generated images (Cesar Cadenas - TechRadar) - This is worth noting, especially because these are tools that don't seem to be currently available in DALL-E 3, which works off the same OpenAI models.
—Need a refresher on Microsoft Copilot? Michael Muchmore has you covered in PCMag’s What Is Copilot? Microsoft's AI Assistant Explained

What is the difference between artificial intelligence and machine learning? (Stephen Perkins - Android Police) - You've likely heard these terms bandied about in a way that feels synonymous. Maybe someone has mentioned them in the same breath as a hedge against using either of the words incorrectly. Or maybe the terms are used correctly, but the difference seems without distinction. Hopefully this explainer will help.

Should You Buy a Vision Pro? A Guide (Dan Shipper - Chain of Thought) - There are still tons of reviews of the Apple Vision Pro out there for you to read, but this boiled things down to the fundamental question. (For me, the answer is the device is about $2k overpriced for how I personally work and consume media.)
—And this post on Reddit, Blind guy tries the Apple Vision Pro, really shows off the unheralded power of this device. Come for the dinosaurs, stay for the ability to see clearly for the first time in your life.
—But what if you have glasses? The AVP is a highly personalized device, right down to your vision prescription. So you aren’t supposed to use glasses. But can you? According to Kyle Riesenbeck in UploadVR, Wearing Apple Vision Pro With Glasses Works, But It's Risky.

YouTube says a Vision Pro app is ‘on the roadmap’ (Nilay Patel - The Verge) - Back when we were experimenting extensively with 360-videos, Nebraska Public Media created 40+ productions. I was really hoping to be able to view them in the AVP, but YouTube is still one of the only ways to distribute 360-videos to a mass audience. So, alas and alack. But this piece gives me hope.

Disney invests $1.5B in Fortnite maker Epic Games to build a persistent game universe (Dean Takahashi - GamesBeat) - I see this investment (a 9% stake) as a big vote on the future of gaming and next-gen internet (aka the metaverse). It reminds me of when Disney invested in Hulu (a 27% stake) back in 2009. But, in fairness to metaverse skeptics, this investment in Epic does show a steep drop in the valuation of the privately-held company, from $31.5B about two years ago (when Sony and Lego invested) to $22.5B now.
—In other Roblox news, as Ryan Morrison asks in Tom’s Guide, Roblox rolls out real-time AI translation for all users — is this the start of true global multiplayer?

And finally…

Vesuvius Challenge 2023 Grand Prize awarded: we can read the first scroll! - And finally, I’m including this both because it's a cool application of AI, and because one of the three students, Luke Farritor, studies at the University of Nebraska-Lincoln.
Bonus: Want to see a project Luke worked on before landing a six-figure prize? Check out this 360-video produced by our radio program director, Genevieve Randall. Luke was the exhibition engineer who worked with keyboardist Paul Barnes to make all those guitars playable in 3D space.

Have a creative, productive week!