Welcome back to another edition of Baekdal/Basic, the newsletter focusing on quick news and trend analysis. Today we have a number of interesting topics to discuss.
The other day, we came across a tweet from Amnesty International that illustrated why journalism and AIs don't mix.
What Amnesty did was to post a series of tweets highlighting the continuing human rights violations by the Columbian police. In itself this is fine... until you look at the image.
This is what they posted (3rd image in a longer series of tweets):
Notice the label at the bottom. It says: "Illustrations produced by artificial intelligence".
Almost immediately, that spurred a ton of negative reactions for people online, telling Amnesty that they should never use AI to create fake images for human rights violations (which seems obvious). Most people said it was wrong to do, others said that it created a trust problem, and others pointed out that when they use fake AI created images, it's easy for pundits or conspiracy nutcases to use that against them.
In other words, it was monumentally stupid for them to do this, and the negative reactions became so vocal that Amnesty had to delete their tweets and the campaign.
Mind you, Amnesty claims they did this to protect the public, arguing that if they had used a real picture, the people in the portrait would potentially be targeted. I get that argument, but still... not the right way to solve this.
However, as a media analyst, I find this fascinating for two reasons:
First, I was fascinated to see how many 'normal people' spoke out against this. I don't have any specific data for this, but it seems to me that the public has had enough of 'fakeness' online. And so using AI in relation to factual reporting will instantly make you less trusted and likable.
And secondly, this also very clearly illustrates why we cannot use AI in the newsroom either.
Mind you, I have no problem with publishers using AI as a design element. For instance, I recently did a presentation for 200 journalists, and while I was talking about the difference between podcasts and newsletters, I needed some kind of graphic to sit behind me. So I turned to MidJourney to create this 'podcast' poster:
But this is not "news" or "newsroom" related. It's just some design, similar to how you might use a royalty free photo.
The problem starts as soon as publishers start to use AI in relation to reporting the news, like Amnesty did. Using a fake-image creator to do that is just an obvious way to destroy people's trust in you.
And for newspapers, where low public trust is a constant, and where we are constantly under attack from people who try to undermine our credibility, using AI created images seems like just the most idiotic thing you can do.
Speaking of illustrations, I came across another example that was much more impressive (thx to Corinne Podger for the heads). ABC News in Australia wanted to report on the coming tax changes ... and the way they did that was just massively impressive.
What they created was a very long scrolling page where, as you go down, you get to see just how massive this tax cut really is, how much could have been paid for it, and who actually gets to benefit from it.
This picture doesn't do it justice, because it's the scrolling that creates the effect. But I wanted to highlight it because it's such a wonderful example of explanatory journalism done right.
There is a new study out from Deloitte, the "Deloitte 2023 Digital media trends", and there is one part of it that is particularly interesting to us as publishers. It's about why people churn.
The data looks like this:
So what are we looking at here? Well, it's looking at services like Netflix, Disney+, Hulu, Amazon Prime TV, etc... and what they found was that 88% have paid for a streaming service, but then things start to diverge.
For the oldest generation, only 24% (which is still a high number) have canceled a streaming service within the past six months, and even fewer, 7%, have resubscribed later.
In comparison, Generation Z and the millennials have a massively different pattern. 62% (almost 2 out of 3 people) have canceled a streaming service within the past six months, but also 43% of them have resubscribed later to the same service.
This is one of the important trends at the moment because it illustrates how the future is about 'switching', rather than 'keeping'.
Think about what is happening here. In the past, people would subscribe to something and then stick around for a very long time. For instance, my father subscribed to one of the national newspapers in my country in his early twenties, and has remained a subscriber ever since. That's 50 years he has subscribed to the same newspaper.
But the younger generation don't think like this anymore. Today, we have so many options, and so many choices, that instead of sticking to one service, we switch it up whenever we want to see something different.
This trend is completely undermining the old subscription model, in pretty much every single way.
First of all, it undermines the concept of a package. People are not subscribing to Netflix. They are subscribing to something specific on Netflix. A specific show or a series of movies that people would like to watch.
But think about this in relation to magazines and newspapers. Today, most magazines are also sold "as a package", but the younger generation don't care about getting a "monthly package of random articles". They are subscribing because of something specific that you do.
It's the same with newspapers. Every newspaper is sold as a package of random news that you get every day, to an audience who don't care about random news, but wish we did something more meaningful and specific.
But also think about what experience this requires. If 62% cancel their subscription within six months, and 43% of them return later ... you need to have a seriously user-friendly and customer-focused cancellation and resubscription process.
Netflix and the other streaming services understand this, and they are doing a good job managing this (well, apart from sounding a bit desperate when you don't).
But think about your magazines and newspapers. Subscribing, changing your subscription plan, canceling, and resubscribing must all be equally easy to do. If you make that difficult, or if you give people an experience that angers them, you will still lose the 62%, but you will never regain the 43%.
This trend can make or break our subscription revenue.
Mind you, right now we are seeing this trend mostly for streaming platforms, but the future outlook for this is that this will be the new normal.
So, look at this graph again. The future is defined by an audience who will switch things around at a much faster rate, but also come back later if you are worthy.
And what defines that is what you put in your newspaper or magazine, not the package itself. It's your articles, your services, and your focuses that will make or break you.
What happens to the future of news if everything just becomes an opinion?
The more we use automated tools, the more important it becomes to also create 'originals'
Everyone is talking about ChatGPT and MidJourney, but their size is also their downside.
Founder, media analyst, author, and publisher. Follow on Twitter
"Thomas Baekdal is one of Scandinavia's most sought-after experts in the digitization of media companies. He has made himself known for his analysis of how digitization has changed the way we consume media."
Swedish business magazine, Resumé