Exploration #93

Hate the Long Scroll? Jump Around!

• If You Click Only One…

• Things to Think About…

• Things to Know About…

• And finally…

Hi all. Welcome to the second of this week’s two explorations (read Exploration #92 here). While the previous exploration caught us up with news and thoughts leading into the new year, this covers all the ground since we returned from the break. From here on in, new explorations will be published once a week on Wednesday.

I’m keeping the focus on generative AI (gAI) for this exploration to align with our 3rd Thursday webinar today at 1pET “Innovate with Current: A Live User’s Guide to Generative AI Tools.” There is still time to register here.

And I’d be remiss if I didn’t give you a chance to join the Public Media Innovators PLC as well.

Regarding gAI, the pace to start the year seems to be matching the pace for much of 2023, and the big announcement this week was the fully productized arrival of Microsoft Copilot for most Windows and Office users (see “If You Click Only One” below). This follows tight on the heals last week of OpenAI’s announcements of the GPT Store and ChatGPT Team (see “Things to Know About” below).

I know some folks think of this as an “AI” newsletter, and that is definitely part of the portfolio. But our focus is really on the emerging media tools and trends that may have the most impact on the future of public media. In future explorations we’ll be broadening our focus to look at next-gen internet (don’t call it the “metaverse”), spatial media/computing (don’t call it “VR/AR”) and game design and development. We may even come back to Web3 (don’t call it “crypto”) at some point as well.

Before I release you to the links below, I want to say thank you to Brian Lorbiecki at PBS Wisconsin. Brian is the designer behind the Public Media Innovators graphic look and feel (logo, header, color palette). Nothing feels quite as good as a great outfit, and Brian dressed us to the 9s.

Okay, on to the links…

If You Click Only One…

Bringing the full power of Copilot to more people and businesses (Yusuf Mehdi - Microsoft Blog) - I don't usually lead with product announcements, but this one feels significant. If generative AI is going to ultimately live up to the hype and become a paradigm shift in how we interact with information, it actually has to fade into the background and become a fundamental part of how we live our lives. This doesn't happen overnight, even for AI. But if anything will hasten us to that future, it’s Microsoft’s inclusion of gAI in its core product for global business.

Time will tell if this is going to be like Windows and Internet Explorer, or if it will be like the Zune. And that will be determined by millions of individual decisions, like the one you’ll have to make. To that end, you should know that the features are largely like ChatGPT Plus and ChatGPT Team, which isn't surprising given that ChatGPT Plus and Copilot share the same underlying model (GPT4-Turbo). Even the pricing is essentially the same (something I suspect antitrust lawyers may eventually find interesting).

Sizing things up, if you live your life around Office365 products, you'll want a paid Copilot subscription over ChatGPT Plus. And, if you are a Teams-centric shop, the more expensive license may be the way to go. But if you switch between Office365 and/or Google and Apple products, ChatGPT Plus probably gets you a slightly more advanced feature set (not to mention access to the latest, advances in the state of the art). And both ChatGPT and Copilot come in a free tier, so you can always try before you buy.

—Extra credit: here's the same announcement from the Office365 perspective (with a hand chart comparing the different pricing tiers and features: Expanding Copilot for Microsoft 365 to businesses of all sizes
—And here is reporting from Tom Warren at The Verge and Kyle Wiggers at TechCrunch, if you'd like to see how the tech press is covering this.

Things to Think About…

ChatGPT and its AI chatbot cousins ruled 2023: 4 essential reads that puncture the hype (Eric Smalley - The Conversation) - A lot of the hopes and fears about AI end up really being directed at the concept of artificial general intelligence (AGI). But if you use regularly use generative AI tools (which, for now, is what we focus on here) you quickly discover their limitations. This piece does a great job of putting those limitations in perspective, and features some good links to additional sources of info for those who want to dig deeper.

Use Of AI In DeepFakes Accelerating Risks To Companies (Cindy Gordon - Forbes) - There is quite a bit of justifiable concern about cybersecurity (especially from threat vectors augmented by AI). However, as this op-ed points out, awareness of deepfakes is an increasingly necessary component of media literacy.
—Here's some additional background from Kevin Collier at NBC News, in case the deepfake concept is somewhat new to you: Deepfake scams have arrived: Fake videos spread on Facebook, TikTok and Youtube
—Then, Parmy Olson at Bloomberg reports on the political angle in Can Midjourney’s CEO Stop a Storm of Fake Election Images? ($)
—And as Paul Sandler reports in Reuters, this can also come through as questionable marketing stunts: AI Elvis to make virtual reality comeback in London show
—But as Mohar Chatterje reports in Politico, when it comes to deepfakes in the wild, we're not ready: A New Kind of AI Copy Can Fully Replicate Famous People. The Law Is Powerless.

AI and Journalism Need Each Other (Francesco Marconi - Wall Street Journal) - The drop line from this op-ed by the Journal's former R&D Chief explains, succinctly, why we in Nebraska are still allowing our website to be crawled by AI bots: "Without the information news organizations gather, artificial intelligence is garbage in, garbage out." Note, this one is behind WSJ's paywall. ($)

Things to Know About…

Introducing ChatGPT Team (OpenAI) - Finally. Claude (from Anthropic) initially launched with a lightweight team structure (though no collaboration) and I've been hoping OpenAI would eventually roll out something similar. It costs more, $25/user/month if paid annually or $30/user/month for a month-to-month subscription (same as the premium team subscription for Microsoft Copilot). The individual ChatGPT Plus version is only $20/month, so why pay more? Privacy. Team accounts are not used as training data.
—Kyle Wiggers reports on the announcement in TechCrunch: OpenAI debuts ChatGPT subscription aimed at small teams

Introducing the GPT Store (OpenAI) - First announced in November, the GPT Store is the first stab by an AI company at creating an app store.
—David Karpf's piece in The Atlantic gives some useful context: ChatGPT’s FarmVille Moment
—Pranav Dixit at Vox asks the right question in: There are too many chatbots
—Also related (and not surprising) is this report from Michelle Cheng at Quartz: AI girlfriend bots are already flooding OpenAI’s GPT store

GPT-4 Turbo now powers Microsoft Copilot. Here’s how to check if you have access (Michael Nuñez - VentureBeat) - From my read, this refers to Windows Copilot, not Copilot for Office365 (which costs $30/month). The check provided by Nuñez is a little techy, but if you follow the instructions it isn't too challenging.
—In other Microsoft news, there's been some buzz about this announcement, so it feels worth mentioning. But I honestly don't see how this is any different than the button I press and hold on my iPhone to raise Siri: Microsoft Swaps Menu Key For Copilot AI Key on Windows Keyboards, Directs OEMs To Do The Same

GitHub Copilot Chat now generally available for organizations and individuals (Shuyin Zhao - Github Blog) - While this would not replace a software developer, it can augment the work of team members that have some dev skills or have taking some coding courses. As we've seen at High Plains Public Radio, some fundamental knowledge can be magnified significantly by these tools.

Getty and Nvidia bring generative AI to stock photos (Emilia David - The Verge) - You may recall last year when Getty rolled out its own gAI tool, after suing Stable Diffusion for using its images as training data. That tool was enterprise-focused and offered up noteworthy indemnification for use. Now Getty brings the generative image generation to smaller companies via iStock. Worth noting: "...the Getty Images AI platform and the new iStock service revolves around legal indemnity. Unlike Generative AI by Getty Images, users will not have unlimited indemnification. The iStock platform will have a cap of $10,000 per asset, the same license it offers for its current library. As with Getty’s first generative AI platform, contributors whose content was used to train the model can participate in a revenue sharing program."

Adobe’s new AI-powered Premiere Pro features eradicate boring audio editing tasks (Jess Weatherbed - The Verge) - While not the generative AI tools we normally discuss here, these 'incidental AI' enhancements may be relevant to the Adobe Premiere users in the community.

Actors union announces deal for AI voice acting licensing in video games (Angela Yang and Chloe Melas - NBC News) - It’s worth noting that this deal isn't with a game studio, but with an AI company. I'm watching this space closely, as we're developing games that will require voice actors.
—Jules Roscoe from Vice brings us a different perspective: 'None of Us Want Our Voices Replicated': Voice Actors Say Union's 'Ethical' AI Deal Is Bad for Humans
—And more reporting from Brock Dumas at FoxBusiness: Video game voice actors furious over SAG-AFTRA union's AI licensing deal 

CES 2024 is not the year of the AI gadget (Max A. Cherney - Reuters) - CES had lots of AI. But it seems, from afar, to be more incidental AI (or machine learning) than generative AI.
—And here's another CES recap, though with more of a TV focus, from indie tech reporter Janko Roettgers: Your 2024 CES smart TV cheat sheet

The First Hit Generative AI Product After ChatGPT? Somehow, It’s Snapchat+ (Alex Kantrowitz - Big Technology) - AI seems to have a knack for reviving companies. ICYMI, last Friday Microsoft overtook Apple as the world's largest company (by market cap), thanks mostly to its sizable bet on AI. AI has helped Meta investors put that whole metaverse thing behind them. And Snapchat seems to be undergoing a resurgence as well.

Amazon’s Alexa gets new generative AI-powered experiences (Lauren Forrestal - TechCrunch) - Remember when voice was all the rage and Alexa was the state of the art for "AI"?

Volkswagen says it’s putting ChatGPT in its cars for ‘enriching conversations’ (Andrew J. Hawkins - The Verge) - On the surface, this feels like it will be less like Knight Rider and more like that scene from The Office when the car GPS guides Michael Scott into a pond. Still, for those thinking about how public media can continue to hold a share of attention in the automobile cabin, this is worth tracking.
—And it’s not just VW: Mercedes-Benz’s best-in-class voice assistant is getting an AI boost
—Or Mercedes Benz: BMW turns inward for CES, announcing new gaming, streaming, AR, and AI features

The best AR glasses of CES 2024 (so far) (Jason England - Tom's Guide) - There's lots of press about Apple's Vision Pro and Meta's Quest 3, but the Xreal Air 2 Ultra appears to be the AR glasses to beat right now.
—More on the Xreal specs from Jason England: Xreal’s new AR glasses bring the fight to Apple Vision Pro — true spatial computing at a fraction of the cost

Apple Announces Vision Pro Release & Pre-order Dates (Ben Lang - Road to VR) - ICYMI, preordering for the Apple headset goes live tomorrow (Friday, 1/19) at 8aET/5aPT. At $3,499 it's pretty spendy, but keep an eye on the hype that follows as Apple moves through its first development year of spatial computing.
—For the photographers (or lovers of photography) in the crowd, tech OG Om Malik weighs in on Why Vision Pro Will Change Photography

Sony Announces XR2+ Gen 2 Headset For Enterprise 'Spatial Content Creation' (Ian Hamilton - UploadVR) - Not to be outdone by Apple, Sony is rolling out a new xR headset not tied to its PlayStation gaming ecosystem. Note the use of the word "spatial." Same as Apple. It this the first post-gAI hype cycle emerging?

Wi-Fi 7 Launches With New Features That Promise Reduced Wireless VR Latency (Ben Lang - Road to VR) - Boring, but important. I know a lot of us probably don't even have WiFi 6 yet, but these standards take a long time to penetrate the market. So it's worth noting that part of the future was written with the adoption of this industry standard.

And finally…

Tropicana rejects 'AI' to become Tropcn (Adam England - Creative Bloq) - And finally...does anyone else here boycott products solely because of inane marketing? <<long-suffering sigh>>

Have a creative, productive week!