The launch of Fast News is terrifying because we’re about to embark on a very public and evolving experiment in AI publishing that we can find no template for here or abroad in B2B – and it could burn fast and out.  

Intriguing because somewhere in there it has the potential to deliver what the AI apologists pledge this rapidly advancing technology can offer: efficiencies and a lighter workload which yields more of better. It’s music for Mi3’s stretched resources and possibly naïve editorial ears.

Finally, there’s some hope because if Fast News lands, it will help the horribly pressured economics of B2B publishers pursuing original, quality journalism and content (most have turned to events for margins).

Sugar and kale

You might have noticed Mi3 is not a sugar hit. More kale, really. For the most part, our content requires focus and effort – quite a contradiction given the levels of attention deficit and information overload we each face daily.

But so far we’ve zigged the zag. And for those who’ve been in their careers at least ten years, Mi3 is the most favoured in part of our competitive set – 4000-plus people working in media and agencies.

The latest NPS scores for trade titles is from our sister company, Media i, which twice every year surveys thousands of media agency and media owner exec views across a range of benchmarks.

That Mi3 is the only industry title by some distance which is in positive territory for content to be shared or recommended is super encouraging. Media and agencies are one part of the broad and complex industry we cover – but it’s an important and bustling sector and we think the sentiment is reflective of the broader marketing supply chain we report on.

However, the chart above, when viewed with below, points to some gaps for Mi3; namely, those who are in their careers less than ten years. We remain the leader in NPS rankings but it’s a much thinner margin.

Who, what, when?

Why? We think partly because there is much interest for those earlier in their careers on people moves and promotions, company announcements and the campaigns and work that this fast-moving sector produces. We call it ‘transactional’ news and although some may beg to differ or dismiss, it is fundamentally human to want to know who is going where and doing what. Until now, Mi3 Editors could only choose one of two pursuits – original coverage or the tonnage that comes with industry news, moves and campaign announcements and media releases.

The mainstream arrival of this crazy thing called generative AI means we can now attempt to do both at the same time. And in doing so, we hope to cover in a new way the huge volume of industry news - but still chase the edge that comes with editors being in-market, working the phones, gathering disparate views and joining dots. 

The idea is for Fast News to allow Mi3 Editors to stay with important industry trends, developments, conversations and debate by using the efficiencies of generative AI and large language models (LLMs) to publish the high-volume news we couldn’t otherwise. Fast News is created deliberately and distinctly separate to the Mi3 flagship. It has its own daily newsletter edition; it occupies its own real estate as a standalone vertical inside Mi3; it has its own LinkedIn news feed and the published stories only appear in Fast News.       

Fast experimentation

It’s an evolving project, as everything is with generative AI, but always with oversight from Mi3 Editors. Nothing is published without Editor approval. And as I said at the top, Fast News is a very public experiment in AI-based B2B publishing – we’ll constantly refine how it’s developed and delivered.

For now, that starts with a few bells and whistles which may have a short life – or not. For instance, we’re putting ChatGPT and Midjourney to the test. ChatGPT is creating the text for the reworked Fast News announcements and media releases. For a bit of fun, it will also generate an automated Haiku for each story we publish. Midjourney, meanwhile, will build images from what comes in with announcements and media statements or from prompts we create.

There are certainly lessons already. Midjourney images with multiple people need more time and better prompts. That’s not fast so for now we’re likely to use supplied images in those instances. We will also vary and experiment with the visual style of our images – that’s part of fast experimentation.

Salesforce and ThinkNewsBrands: kudos

Finally, a big shout-out to our launch partners, Salesforce and ThinkNewsBrands. We’ve been drilled and grilled by both and rightly so. But both partners understand applied AI is a critical part of the industry’s future and this is one way for those in marketing, customer-related tech, media and agencies and advisors to watch and learn. Although the Fast News project is no charity for either ThinkNewsBrands CEO Vanessa Lyons or Salesforce APAC CMO Leandro Perez, kudos to them and their teams for backing the adventure.

We hope you join us for the ride and some lessons too. Fast News won’t be perfect, nor will it be set and forget. This technology is evolving with velocity and somehow our small Mi3 team has got to hang on and keep pace. That’s the intent. Send us your thoughts now and later. Who knows what Fast News will look like in a year. 

Here we go…

Paul McIntyre
Executive Editor 

Don't forget to: 

• Follow Fast News on LinkedIn here
• Send your media announcements to fastnews@ai.mi-3.com.au