AI came up time and time again at last week’s INMA World Congress of News Media in NYC. It was the top concern of the CEOs and was mentioned in almost every session.
Let’s break it down:
Applications that can be right used now
Karen Silverman of the Cantellus group highlighted the following: news aggregation, transcription, drafts of articles, discovery and idea generation, investigation of known subjects, investigating novel patterns in structured and unstructured data, immersive storytelling (and graphic generation), restoring footage, targeting readers, moderation, UGC, translation.
That’s a lot of efficiency.
The copyright issue came up, led by Robert Thomson, Global CEO of News Corp. It’s a big one and hard because LLM (Large Language Models) are, at least in part, trained on our material.
How much? I am sure there is data out there, but I am not sure we truly know. And I am fairly sure that it’s just a fraction of the content out there.
Here’s my two cents, which I know isn’t popular: I think the ship has sailed and it will likely end up being a huge distraction to fight. I hope there are agreements that can be reached for specific quotations and charts.
But frankly, we also save to look inwardly. If I write a restaurant review and speak to 10 diners and four staff at the restaurant, two of which I quote — whom should I pay? The people I quoted? The people I spoke to that helped form my opinion? In our case it’s no one. My point is the conversation is so nuanced that it’s hard to find a solution that works across the board.
Scraping media content
There will be a tipping point where SEO doesn’t give us enough return to warrant the scraping of our content. It’s then that we decide to close off our content to search, meaning LLMs are no longer trained on our information and the value to our customers is that of verified independent truth.
What does that mean to our top of funnel? Where does it come from if not search? I wish I had the answer to that.
Is there an existential threat of AI?
Yes, I am convinced a lot of people will end up with their preferred chatbot type service that meets the majority of their needs — and that does not bode well for news generally. But I left the event a lot more confident about the future of news and the value of real humans overseeing our news.
We need to continue to work on trust and independence, but maybe that is precisely where the opportunity is.
If you’d like to subscribe to my bi-weekly newsletter, INMA members can do so here.