Context Window 44
This week’s stories point to deeper layers of the AI stack: standards, infrastructure and how organisations are evolving. From licensing to litigation to content production, there are clear signs of where publishing may be heading.
A significant licensing development this week, as a group of large online publishers including Reddit and Yahoo came together to support the launch of RSL (Really Simple Licensing), a new technical standard to communicate machine-readable AI licensing terms for online content, and manage licensing through a collective protocol. Book and journal publishers O’Reilly Media and MIT Press are involved. It remains to be seen whether the big AI platforms will support the initiative, how compliance can be monitored and enforced, and how this will sit alongside direct licensing deals. However, from a policy perspective, the more robust and straightforward options exist, the harder it is for tech companies to argue that copyright exceptions are necessary because licensing isn’t practical.
Shifting gear to legal arguments over previous AI training, the landmark class action settlement in Bartz v. Anthropic hit a buffer with Judge William Alsup declining to approve the settlement and asking both sides for more information. Despite some of the more simplistic analysis last week, the path to resolution will be slow, complex, and contested.
Meanwhile, new litigation was filed this week by publishers Merriam Webster and Encyclopedia Britannica against Perplexity, and book authors against Apple…
Anthropic announced a significant upgrade for Claude, which can now write Word, Excel, PowerPoint and PDF files as outputs. This makes it considerably more useful for a lot of general business tasks. It comes as Microsoft announces that it is integrating Anthropic AI into Office 365, alongside its longstanding partnership with OpenAI.
Google has partnered with non-profit educational publisher OpenStax to turn six of its existing textbooks into interactive experiences for its NotebookLM model. This follows earlier partnerships with news publishers such as The Economist and The Atlantic. The NotebookLM model of high quality, tightly curated source material combined with AI analysis and outputs is a really interesting one: if I were a commercial textbook publisher, I would be all over this and custom GPTs for my titles.
There’s an interesting piece in HBR on how AI is disrupting the traditional consulting business model, which historically relied on a stable, pyramid model: a broad base of junior researchers and analysts supporting successively smaller tiers of senior consultants and management. The base tier is the one most disrupted by AI, and threatens to topple the entire edifice. All interesting in the abstract—but this matters to all of you because the model described is one that is common to knowledge-based businesses, and there are clear parallels with publishing. The new organisation model proposed in the piece—a narrower, flatter company—looks like many of the publishing startups I’ve seen in the last year, and preserves a pipeline of future senior talent, a key consideration for future proofing the org. Whichever tier you’re currently at, you will need to understand the implications of this shift.
If that piece provided a challenging but essentially hopeful perspective on AI and publishing organisations, I have to admit that I found this one more depressing: an AI-powered podcast studio with a content team of four people, producing 3,000 episodes a week, each costing less than $1 and an hour of production time, which means profitability on as few as twenty listeners per episode. The thesis behind this: “We believe that in the near future half the people on the planet will be AI, and we are the company that’s bringing those people to life.” There are real challenges to the podcast business model, but I deplore this as an answer. Optimising only for volume breaks any reasonable ratio for signal to noise, and makes discoverability a more acute challenge for every other publisher on distribution platforms—in the same way that the Kindle store remains an amazing convenience if you know exactly what you’re looking for, and for me at least, almost unusable for browsing and discovery. For thoughtful publishers, this underlines the importance of building loyal audiences and differentiated voices.
Finally, an admin point: a couple of readers have reported that their email software blocks hyperlinks in this newsletter. If that is happening for you, please would you let me know what software you’re using, so I can try to get to the bottom of it? This was originally published in my email newsletter. To receive weekly updates on how AI is affecting the publishing industry, sign up here.