The mushy middle of AI adoption

In short

This episode discusses the current state and future of AI in the media industry, focusing on how AI can improve news products rather than just production efficiency. Guests highlight that despite the hype, AI's practical impact on newsrooms is still limited. They share experiences with AI experiments, noting both successes and challenges. The future of news companies is explored, with potential shifts towards becoming API providers and the increasing importance of quality journalism and human connection. The conversation also touches on the value of information in the AI era and the need for effective mechanisms to realize that value.

Key takeaways

• AI's primary impact on media is currently focused on production efficiency, with less attention on improving the news product itself.
• Despite significant AI advancements, newsrooms have not experienced drastic changes in their daily operations compared to previous years.
• The future of news companies may involve a shift towards becoming API providers, emphasizing quality journalism and human connection.
• Quality information is becoming increasingly valuable in the AI era, posing a challenge for news organizations to effectively monetize it.
• Media companies need to move beyond incremental AI adoption and consider more fundamental changes to their business models and products.

Chapters

1. 00:00

   Introduction to Beehive

   The episode begins by introducing Beehive, a platform designed for media businesses, particularly those scaling newsletters. It's highlighted that email is often the backbone of such businesses, and Beehive aims to simplify its management.

2. 00:37

   Beehive's Origins and Users

   Beehive was built by individuals who scaled Morning Brew and The Information. It's designed for serious media operators, not as an afterthought for marketers. Companies like TechCrunch and The Texas Tribune already use it for their newsletters.

3. 01:15

   The AI Discussion in Media

   The conversation shifts to AI in media, noting that most discussions focus on distribution threats and production efficiencies, rather than how AI can improve the news product itself.

4. 01:45

   Martin Scurr's Background

   Martin Scurr, a guest on the show, has experience from both large media organizations (like Aftonbladet) and startups. He led a dedicated AI hub at Aftonbladet and recently co-founded Hit Media, a newsletter startup targeting young women.

5. 02:30

   Current State of AI in News Production

   Scurr describes the current state of AI integration in news production as 'nibbling around the edges' and argumentative, with few tangible, impactful use cases emerging since the ChatGPT moment.

6. 03:10

   AI's Impact on Newsrooms

   Despite the hype around AI, Scurr notes that newsrooms haven't changed drastically compared to five years ago, suggesting that the practical impact on daily journalistic work is still limited.

7. 04:00

   Aftonbladet's AI Experiments

   Scurr details Aftonbladet's AI initiatives, including an AI hub focused on developing tools for the newsroom. They experimented with various tools, including chatbots and text-to-speech, over a year or two.

8. 05:00

   Challenges and Successes of AI Tools

   While some AI tools, like transcription, were successful, many editorial tools aimed at efficiency gains were not. Scurr attributes this partly to the technology not being ready and partly to a focus on the tech itself rather than solving specific problems.

9. 06:00

   The Future of News Companies

   The discussion explores how news companies might evolve, potentially becoming API providers rather than direct-to-consumer businesses, focusing on quality journalism and human connection.

10. 07:00

    AI and the Value of Information

    The conversation touches on the increasing value of quality information as AI models need fresh data, and the challenge for news companies will be realizing this value through effective mechanisms.

Get the next The Rebooting Show recap

A short recap of every new episode, by email. Free for up to 5 shows.