Exploration #92

Table of Contents

• If You Click Only One…

• Things to Think About…

• Things to Know About…

• And finally…

Welcome to the new, Beehiiv-based version of the newsletter formerly known as Public Media Labs. New year, new platform, new name. Thanks to all of you who have followed me out of the Outlook wilderness, and welcome to all of you who signed up over the holidays, courtesy of NETA’s announcement of the Public Media Innovators Peer Learning Community (PLC). The next few editions will probably involve a bit of trial and error as I adapt for Beehiiv some of the design choices I employed in Outlook, and then adapted to Substack.

What happened to Substack?

So…about Substack. I was planning to publish this newsletter last week but missed the mark after ditching my previous plans to use Substack. If you haven’t followed their content-moderation-meets-Nazis drama, you can catch up with The Atlantic here, and Platformer here, here and here. The TL;DR is that Substack’s attempt to evolve into a social network gives them incentive to push “popular” newsletters regardless of whether they are quality tech blogs, quality freelance journalists, or Nazis.

For now, I do still plan to feature writers who use Substack. Each independent author needs to make their own choice on this issue based on their own personal situation and resources. I’m not naive enough to think that fascists can’t also use Beehiiv (or any Substack competitor). But for this newsletter, migrating into a platform that, at best, has an aggressively casual attitude toward profiting off fascist ideologies creates associations I do not want inadvertently applied to me, the Public Media Innovators PLC, NETA, or Nebraska Public Media.

Before moving on, I do want to quickly say thank you to all those folks who helped me think this through over the last couple of weeks. You know who you are.

“3rd Thursday” Webinar This Week

Now on to happier topics. We have officially launched the Public Media Innovators PLC. If you are getting this newsletter but haven’t joined the PLC yet, you can sign up here.

We’re also kicking off our new “3rd Thursdays” speakers series this week. On the 3rd Thursday of each month at 1pET/10aPT, we’ll host a Zoom webinar with a speaker or panel discussing an aspect of emerging media as it relates to public media. The first one “Innovate with Current: A Live User’s Guide to Generative AI Tools” will happen January 18 at 1pET. Register here.

Two ‘Explorations’ to Catch Us Up

One last note: we’ve got a lot to catch up with since the last Outlook newsletter published on December 18. To make up for not publishing last week, I’ll be publishing two Explorations this week. This one catches us up to the start of the year. Lots of predictions for 2024 in this one.

The next Exploration will cover the news from the New Year forward (including OpenAI’s GPT Store and ChatGPT Team). That will drop Thursday morning. After that, I plan to publish new Explorations each Wednesday.

Okay, on to the links…

If You Click Only One…

Michigan Radio project points to future of AI use in newsrooms (Dustin Dwyer - Current) - ICYMI, this dropped via Current just prior to Christmas. Dwyer’s experience at Michigan Radio is a great example of how people around the PBS system and NPR network are beginning to use generative AI to advance their mission by doing more with the same resources.
—Extra Credit: This piece from Alex Fregger at High Plains Public Radio also dropped in Current last month: How one station used AI to automate promotions and save time

Things to Think About…

Predictions for Journalism, 2024: AI gets widely adopted by smaller newsrooms (Ernest Kung - NiemanLab) - Part of NiemanLab's predictions for 2024, it's not hard to see the connection between Dustin Dwyer's piece for Current and this. Below are some other NiemanLab 2024 predictions of note:
—The AI copyright watershed comes into view by Richard Tofel
—The press adopts a new level of transparency around images by Ståle Grut
—Raising the alarm bell for public media by Kerri Hoffman

Gen AI Will Transform News. Experts Say The Rulebook Must Be Written Now (Michael Stahl - TV News Check) - H/t to Liz Maestri at PMVG for this story. I especially liked this sentiment from the BBC: "Ellis went so far as to say that the BBC is 'a technology organization as well as a content organization,' one that is not 'cutting humans out of the loop.' However, she also noted that the company is not blind to the potential pitfalls of AI. 'We’re now trying to work out how all these new capabilities [help us] can create new things, new products for our audience, and how we can do that ethically and safely, because if we lose trust with our audiences, we lose everything,” Ellis said. “Generative AI has its moments and has its foibles, so [we’re] just trying to create a cross organization conversation about where we want to go with it, what we want to do with it, and how we do that safely.'”
—Related: Amy S. Mitchell’s piece in Tech Policy Press: Journalism Meets AI: Shaping the Future with Smart Policies

How do we know when an AI model is good enough? (Yennie Jun - Art Fish Intelligence) - The contributions that Jun makes body of work on biases in AI is valuable, a this summary of her 2023 work shows.

Signs and Portents (Ethan Mollick - One Useful Thing) - If you are new to the newsletter, you're about to learn why I love Mollick's writing. His approach to gAI is values-based, and very much aligns with public media's values. I guarantee this won't be the last time you hear from him, but for now, here is his look-ahead at 2024.
—In a similar vein is this PDF deck from Deloitte entitled Generative AI and the Future of Work

AI is here – and everywhere: 3 AI researchers look to the challenges ahead in 2024 (Anjana Susarla, Casey Fiesler, Kentaro Toyama - The Conversation) - As with Mollick above, I tend to favor the perspective of those in academia. Note, even though they talk about “AI,” they are really referring to generative AI.

What to Expect in AI in 2024 (Shana Lynch - Stanford HAI) - The view from seven faculty and senior fellows at Stanford's Institute of Human-Centered Artificial Intelligence.

What to Expect From AI in 2024 (Will Henshall - Time) - The section about "The Coming Trough of Disillusionment" really resonated with me.

Big Ideas in Tech 2024 (Andreessen Horowitz) - Honestly, I'm not gaga for A16Z the way some folks are. I'm just not philosophically on their page. But I did appreciate the perspective in their Games chapter. Unfortunately, I can't link straight to it, but you can jump to it quick once the table of contents comes up.

Reality Check: Looking Back at Augmented Reality in 2023 (Tom Emrich via LinkedIn) - Emrich writes these look backs every year and I find them to be a good state reflection on the state of the art.
—Related: Emerich’s 24 Augmented Reality Trends to Keep an Eye on for 2024

Things to Know About…

Top Ranked Gen AI Tools Entering 2024 (Routz.Tech) - I'm including this interactive graphic more as a way to understand the multitude of companies and diversity of categories the gAI space than as an endorsement of their specific data and rankings. We discuss the biggest movers and shakers here, but this iceberg is wide and deep.
—You can dig more deeply into that information here.
—And while not as extensive, Keyede Erinfolami gives us a look from a different perspective via Slashgear: The Year Of AI: Here Are The Most Used AI Tools Of 2023

Google Gemini is not even as good as GPT-3.5 Turbo, researchers find (Carl Franzen - VentureBeat) - While I have actually found Bard (with Gemini) and huge improvement (especially, as I've mentioned here, with image recognition), and often use it now when I don't feel like dealing with the fact that ChatGPT Plus and Claude 2 have logged me out (again). Last month a group of researchers put Bard to the test and found it wanting. As they say, you mileage may vary.

Microsoft Copilot app is a stealthy AI launch that you should pay attention to (Michael Nuñez - VentureBeat) - For those who like to access gAI tools on their mobile devices, this could be of interest. Over the holidays, Microsoft rolled out their own gAI app (powered by ChatGPT) for iOS and Android. Depending on how you use gAI, it could be worth some play time. Fair warning, if you aren’t logged in via your Microsoft credentials, you only get 5 interactions a day.
—Of course, you've got to wonder how long till Microsoft starts canibalizing ChatGPT, but Sarah Perez reports in TechCrunch that this hasn't been an issue so far: Despite free access to GPT-4, Microsoft’s Copilot app hasn’t impacted ChatGPT installs or revenue

Anthropic will help users if they get sued for copyright infringement. (Emilia David - The Verge) - While I think a lot of this is optics, and no one has really put this to the test yet, it's still worth noting that at this point companies are at least going through the motions of providing a shield to creators.
—And you can read the announcement from Anthropic here: Expanded legal protections and improvements to our API

The New York Times Wants ChatGPT Gone. Nice Try (Matthew S. Smith - IEEE Spectrum) - Speaking of legal action (see above), between Christmas and New Years, the NYT sued OpenAI and Microsoft. Feels like posturing for a settlement with preferred treatment, if you ask me.

Midjourney V6 is here with in-image text and completely overhauled prompting (Carl Franzen - VentureBeat) The arms race amongst generative art tools didn't slow down over the holidays. With DALL-E 3 (from OpenAI) and others beginning to catch up in terms of quality, the folks at Midjourney released their latest version. Of biggest interest to me is the fact that the sensitivity of the v6 model means users will need to relearn prompting for Midjourney. That seems like a gamble. Given the way other generative art tools are being integrated into other experiences (like DALL-E 3 in ChatGPT Plus), it seems like Midjourney is almost trying to create a professional class to separate generative artist from 'dabblers.' We are still waiting for it to go into wide release via web browsers though.
—Daniel Nest with Why Try AI provides an image-rich look at the upgrades: Midjourney Version 6: Look Ma, Text!
—Also, Rory Flynn ran some 1:1 tests comparing Midjourney v6 to DALL-E 3 and posted them to his Twitter/X account.

Camera Companies Fight AI-Generated Images With 'Verify' Watermark Tech (Emily Price - PC Magazine) - As as photographer, I'd been wondering why major camera companies like Canon, Nikon and Sony hadn't been more active in the gAI discussion (especially in terms of advocating for the creators that invest in their systems). So I'm glad to see some semblance of unity from them on this issue, but I'm concerned this feature will only be applied to new camera bodies. Photographers invest and then hold onto their gear for years, so I hope this eventually becomes a firmware update that can benefit a majority of photographers.

Shapiro Administration and OpenAI Launch First-in-the-Nation Generative AI Pilot for Commonwealth Employees (State of Pennsylvania) - Whether its ultimately deemed a success or not, it's a milestone in the adoption of gAI. Pennsylvania readers especially, take note.
—Related: Sharon Goldman reports in VentureBeat that meanwhile, just across the Delaware, NJ is making its own moves: New Jersey touts new East Coast AI hub at Princeton University 

Apple explores AI deals with news publishers - New York Times (Reuters) - Because the NYT is paywalled, I'm linking to Reuters report on the report.
—And here is a complementary report on the report from Jay Peters at The Verge: Apple reportedly wants to use the news to help train its AI models

Digital inaccessibility: Blind and low-vision people have powerful technology but still face barriers to the digital world (Michele McDonnall - The Conversation) - It's easy to think that accommodating assistive tech is making digital experiences accessible, but usability is about design more than tech. Here's a key line: "The most significant challenges related to assistive technology centered on the inaccessible digital environment: documents, software, websites, graphics and photos. Digital content is sometimes technically accessible but unusable by people who use assistive technology."

Facebook, Google data centers among latest developments transforming Nebraska farmland (Yanqi Xu - Flatwater Free Press) - Interesting developments (literally) here in Nebraska. This is a good reminder that all the ‘magic’ we get living in a world with cloud computing comes at a cost. Is this happening in your state?
—If that intrigues you, you’ll also be interested in this story by Natalia Alamdari from January 2023: Crypto needs cheap energy. It found a whole bunch in Nebraska.

And finally…

Submarine Cable Map 2023 (TeleGeography) - And finally, the latest edition of the global submarine cable map is out. I love that they're allowed to have actual names, like “Bifrost” which connects Singapore and the US. If you love maps, you're welcome.

Have a creative, productive week!