Good afternoon!
The Medium delivers in-depth analyses of the media marketplace’s transformation as creators, tech companies and 10 million emerging advertisers revolutionize the business models for “premium content”.
For my final mailing of 2023, I sent out my Top Essays of 2023 and revisited my Predictions for 2023.
[Author's Note: Last night, I returned from Munich, Germany where I attended The DLD Conference. I was there for two purposes: To attend the conference, as I have for most of the past 12 years since 2012, and to present to a group of trainees and journalists at Burda Media for the second year in a row (here is a summary of last year’s presentation).
The topic of my talk was based on last year’s essay “The Daisy or The Flywheel?” and the Medium Shift column, “A Hopeful Sign for Big Media?”. The talk started with the business logic and economics of the newsletter model of successful independent journalists. I used those examples as the foundation of how to better understand the new “Daisy” or bundled business model at The New York Times. What are the best business models for journalism that can serve customer’s needs in a retail-first, consumer-first marketplace?
If you are interested in learning more about the presentation or would like for me to deliver a version of it tailored to you and your team, you can respond to this email.
From the perspective of a media analyst, the timing of the conference was interesting. Within DLD, many of the talks focused on the various implications of artificial intelligence (AI) across the sciences, the media business (gaming) and the arts. Outside of the walls of the DLD Conference, more than a dozen major corporations across technology, finance and media announced major job cuts.
Are the two connected? In most cases, the answer is no. Only in the instances of Google—which The Information reported last month is laying off a percentage of its advertising sales workforce at YouTube—and language-learning app Duolingo—which laid off 10% of its contractors while using Generative AI to create more content—could any impact of AI on the workplace be discerned. Elsewhere at companies like Amazon’s MGM Studios, Unity Software (gaming) and Universal Music Group, humans are losing jobs in the media business because of cost-cutting unrelated to the emergence of AI.
Much of the hand-wringing about AI in recent years—and especially during last year’s writer’s and actor’s strikes in Hollywood—have been the fears of what AI can do that is similar to or even better than most humans. The news from YouTube and Duolingo is certainly fodder for those fears, and there is the expectation that we will see more stories like them in 2024 and beyond that humans inevitably will be replaced by machines.
Howeverm an interview at DLD with Cassie Kozyrkov of Data Scientific—she was the first Chief Decision Scientist at Google—framed the problem as still a fundamentally human one. She argued in her interview with Axios' Ina Fried that we are “too obsessed with being data-driven”. In order to be “data-driven you have to approach your decision in a particular way”. Otherwise, one is “data-inspired” which she analogized to “a whale swimming around plankton” selecting numbers that were near a decision but did not drive a decision.
Key Takeaway
We are very much in the early days of AI changing media and entertainment and that is more a reflection of the humans building the AI than the capabilities of the AI itself.
Total words: 900
Total time reading: 4 minutes
She then argued any AI system is not based on data, alone. but rather three very important decisions made behind-scenes by humans (which she calls the “Kozyrkov criteria”):
What was the objective?
What were the data (provenance and schema)? What were the examples used for programming (because AI is programming with examples?)
How was the AI tested? What were the criteria for launch?
If we have answers to these three questions, we “know an awful lot about the system” because these are three “highly subjective decisions made by some human or group of humans in a room.” In other words, “Did builders understand those questions?
The AI we see today is only as good as the humans who built it. A message like Kozyrov’s tells us that we are very much in the early days of AI changing media and entertainment and that is more a reflection of the humans building the AI than the capabilities of the AI itself.
Three Insights Into Gaming from DLD
It is also worth highlighting two talks on gaming that I was able to attend and are both now up on YouTube.
First, Hilary Mason, founder of AI-gaming startup Hidden Door, is one of those humans building AI for gaming. Hidden Door is a game technology studio building the first narrative AI — a platform that transforms any work of fiction into an infinite social roleplaying experience, bringing together players, authors and other creators. Meaning, any author can upload their novel onto Hidden Door and, within “only a few hours”, Hidden Door can bring that world to life for fans. Mason sold it as “a Netflix for interactive fan experiences”.
Longtime readers of The Medium will remember I often quote former WarnerMedia CEO Jason Kilar about the overlap between gaming and streaming. As media habits evolve with Generation Z and Generation Alpha towards a mix of streaming and gaming, technology drives the consumer away from formats and more towards Kilar’s argument that “beloved characters and worlds matter.”
Hidden Door is certainly proof in the pudding of that worldview. Its target customer is “somebody who reads a book or watches a movie and just keeps thinking about it.” The platform provides those customers the ability to engage with their “beloved characters and worlds” in ways that are “totally up to” the users and their friends. The platform has more in common with role-playing games like Dungeons & Dragons, and the “infinite” aspect of the platform is that consumers can invent and imagine their own stories.
The offering contrasts starkly with Netflix Games, which seeks to take advantage of fandom for its IP (and others, like Paramount’s Teenage Mutant Ninja Turtles) via mobile games “without ads, in-app purchases, or extra fees.” Gaming studios both acquired by or partnering with Netflix imagine new mobile games around their IP.
Hidden Door’s is fan-imagined, fan-driven and faster-to-market because of AI. Questions around Netflix’s gaming strategy are starting to grow, as The Wall Street Journal’s Jessica Toonkel reported earlier this month, and Hidden Door seems poised to raise more of those questions. I also spoke with a senior executive at one of the European gaming studio partners for Netflix. He noted that community was a key offering missing from Netflix games, but he was bullish on Netflix’s gaming roadmap and believes we will see community and other common features in gaming, soon.
Last, I highly recommend this talk on The Age of AI Hardware, Smartglasses & Spatial Computing from tech and gaming executive Cathy Hackl. It is an excellent encapsulation of how spatial computing is emerging as a new medium and why gaming is most valuable for understanding it. Keep your ears open for a shout-out to my last Medium Shift column—“Media Executives Covet Games, but Are Ill-Suited to Run Them”—towards the end.