Artificial intelligence continues to be one of the major themes of the year after ChatGPT became the application of the technology that made it accessible and its potential understandable for consumers and business leaders alike. Peter Bale of INMA says that AI search will provide a new challenge for publishers because they will struggle to attribute content. We’re now definitely at the innovation trigger of Gartner’s Hype Cycle.
As if to prove Peter’s point, publishers are already lawyering up to challenge generative AI providers for using their content to train AI. In the current regulatory environment, I can imagine that licencing knowledge for the training of large language models could become a lucrative revenue stream for publishers, especially those like Dow Jones that produce a lot of valuable business intelligence. Regulators are much more likely to compel tech companies to share their wealth with content companies.
One of the themes of the past several years has been newsletter growth, and the Post and Courier in the US has three lessons from its Google News Initiative newsletter project. One that leaped out at me was how they focused on ARPU, which actually helped reduce churn by not setting up a meager introductory price.
This plus a few bits of social media news including podcast management tools testing in YouTube Studio. Something that makes my blood boil a bit is that Meta is testing a verified account service. I don’t think I should be paying for Meta policing copycat accounts. That seems to me to be their responsibility. And a story on the ad side about how supply-side platforms have become a low-cost commodity service.
Peter not only identifies one of the challenges that AI search will pose, namely attribution, but his post is a good overview of how that is being handled by different providers including Microsoft’s Bing and a service called Perplexity.ai.
I think that Peter is right when he says that Microsoft isn’t targeting the profits of publishers with its AI search but rather those of Google. But we’re definitely in a disruptive phase of AI innovation, and it’s during these times when potential winners can quickly rise and then fall to become a footnote to history.
Attribution of content and sources — already seen as a big problem by publishing content creators — is set to become a battleground between publishers, companies promoting AI-driven search, and regulators and politicians who generally lag in responding to technology upheaval.
This Bloomberg story proves Peter’s point that attribution and compensation are going to become very important for all kinds of generative AI as the companies behind them seek new sources of knowledge to train their LLMs.
Social Media and Industry news
YouTube has become a major podcasting platform, and podcasters have been uploading simple videos that remind me of a lot of studio cam footage. So it comes as no surprise that YouTube is testing tools for this market.
Meta hops on the verification bandwagon. I think they need to fix the cloning of accounts on their own as an act of good faith rather than charging for this privilege. This is affecting not just brands but regular people. It’s not a good look.
And lastly, as I mentioned last week, with the economy going through a soft spot in the US and several places really, the ad market is shifting quite a bit.